home *** CD-ROM | disk | FTP | other *** search
/ 500 MB Nyheder Direkte fra Internet 2 / 500 MB nyheder direkte fra internet CD 2.iso / start / data / text / personal.txt < prev    next >
Text File  |  1994-09-21  |  133KB  |  2,705 lines

  1.  
  2.           ANALYZE YOURSELF
  3.   The computer can analyze your body and mind.
  4.  
  5.                 Death
  6.   Terrence Lukas wrote a program that predicts when you'll die. 
  7. The program makes the computer ask for your age and sex; then it 
  8. asks about the life and health of your parents and grandparents, 
  9. your weight, your personal habits (smoking, drinking, exercise, 
  10. and sleep), your history of medical check-ups, your social class 
  11. (your education, occupation, and income), and your lifestyle: 
  12. urban or rural, single or married, aggressive or passive, and 
  13. whether you use seat belts. The computer combines all that 
  14. information, to tell you when you'll probably die.
  15.   The program uses the latest statistics from life-insurance 
  16. companies and from medical research. Lukas wrote the program at 
  17. the University of Illinois Medical Center.
  18.   Running the program is fun. Each time you answer a question, 
  19. the computer tells you how your answer affects its prediction. 
  20. You see its prediction bob up and down, until the questions 
  21. finally end, and the computer gives you its final prediction of 
  22. when you'll die. It's like watching the early returns of a 
  23. Presidential election, except the topic is you!
  24.   The computer pops out with surprising comments, based on the 
  25. latest medical research. Here are some comments the computer 
  26. prints:
  27. Professionals usually live longer, except musicians, architects, 
  28. and pharmacists. Why this is true is unknown.
  29.  
  30. Cooks, chefs, bakers, and other people who work at jobs 
  31. associated with overeating have a lower life expectancy.
  32.  
  33. Adults that sleep too much use too many hours in nonphysical 
  34. activity. They may be unhappy and sleep as an escape, or may be 
  35. ill. Depressed people have shorter life expectancies.
  36.  
  37. Moderate drinking (up to two drinks per day) reduces stress and 
  38. aids digestion. Heavy drinking, however, produces physiological 
  39. damage. As for teetotalers, they may have rather rigid value 
  40. systems and may undergo stress in maintaining them.
  41.   The program is on pages 34-36 of the November 1977 issue of 
  42. Kilobaud Microcomputing Magazine.
  43.                                                     Brainwaves
  44.                                          A computer has been 
  45. programmed to read your mind, by analyzing your brainwaves. A 
  46. newspaper article described the program dramatically: you're an 
  47. airplane pilot . . . your plane is going to crash . . . but you 
  48. think ``Up!'', and the plane automatically goes back up! . . . 
  49. because the plane is run by a computer that's reading your 
  50. brainwaves!
  51.                                          But what if the pilot is 
  52. sadistic, and thinks ``Down''?
  53.                                          Anyway, the program 
  54. isn't perfected yet. When the computer tries to distinguish 
  55. brainwaves that mean up from brainwaves that mean down, it gets 
  56. the right answer 75% of the time. In other words, 25% of the time 
  57. it goofs. I'd hate to be in a plane controlled by a computer that 
  58. was having a bad day!
  59.                                          Recently, the U.S. 
  60. Government has shifted the emphasis of this research. The new 
  61. emphasis is more practical: to make the computer ring a buzzer, 
  62. when the pilot's brainwaves indicate the pilot is daydreaming.
  63.  
  64.                                                     Horoscopes
  65.                                          If you hate horoscopes 
  66. as much as I do, you'll love Adrian Thornton's program. It makes 
  67. the computer print fake horoscopes. The computer asks when you 
  68. were born, computes your sign, and then prints an enchanting 
  69. remark.
  70.                                          For example, if your 
  71. sign turns out to be Capricorn, the computer will print:
  72. You are conservative and afraid of taking risks. You don't do 
  73. much of anything and are lazy. There has never been a Capricorn 
  74. of any importance. Capricorns should avoid standing still too 
  75. long, as they tend to take root and become trees.
  76.                                          If you're a Virgo, the 
  77. computer will print:
  78. You are the logical type and hate disorder. This nitpicking is 
  79. sickening to your friends. You are cold and unemotional and 
  80. sometimes fall asleep while making love. Virgos make good bus 
  81. drivers.
  82.                                          You'll enjoy seeing your 
  83. own ``horror''scope. The complete program is on pages 48-49 of 
  84. the February 1979 issue of Kilobaud Microcomputing Magazine.
  85.             Psychotherapy
  86.   In 1966, Joseph Weizenbaum programmed the computer to act as a 
  87. psychotherapist. On the keyboard, you type whatever thoughts come 
  88. to your mind, and the computer will give you therapy. When typing 
  89. on the terminal, feel free to use whatever words you wish ___ you 
  90. don't have to use a ``computer language''.
  91.   The most amazing thing about the program is that it does a 
  92. decent job of therapy, even though the computer doesn't know what 
  93. you're talking about! Weizenbaum's program contains the 
  94. definitions of only a handful of words; but if you use words the 
  95. computer doesn't know, the computer does an amazingly good job of 
  96. ``faking''. For example, if you say, ``My junkapoo is like my 
  97. iliabop'', the computer will continue the conversation by asking 
  98. you, ``In what way?'' even though the computer doesn't know what 
  99. a junkapoo or an iliabop is.
  100.   Here's a conversation between his computer and a patient.
  101. Patient:Men are all alike
  102. Computer:In what way?
  103.  
  104. Patient:They're always bugging us about something or other.
  105. Computer:Can you think of a specific example?
  106.  
  107. Patient:Well, my boyfriend made me come here.
  108. Computer:Your boyfriend made you come here.
  109.  
  110. Patient:He says I'm depressed much of the time.
  111. Computer:I am sorry to hear you are depressed.
  112.  
  113. Patient:It's true. I am unhappy.
  114. Computer:Do you think coming here will help you not to be 
  115. unhappy?
  116.  
  117. Patient:I need some help, that much seems certain.
  118. Computer:What would it mean to you if you got some help?
  119.  
  120. Patient:Perhaps I could learn to get along with my mother.
  121. Computer:Tell me more about your family.
  122.  
  123. Patient:My mother takes care of me.
  124. Computer:Who else in your family takes care of you?
  125.  
  126. Patient:My father.
  127. Computer:Your father.
  128.  
  129. Patient:You are like my father in some ways.
  130. Computer:What resemblance do you see?
  131.  
  132. Patient:You are not very aggressive, but I think you don't want 
  133.       me to notice that.
  134. Computer:What makes you think I am not very aggressive?
  135.  
  136. Patient:You don't argue with me.
  137. Computer:Why do you think I don't argue with you?
  138.  
  139. Patient:You are afraid of me.
  140. Computer:Does it please you to believe I am afraid of you?
  141.  
  142. Patient:My father is afraid of everybody.
  143. Computer:What else comes to mind when you think of your father?
  144.  
  145. Patient:Bullies.
  146. Computer:Does that have anything to do with the fact that your 
  147.       boyfriend made you come here?
  148.   Like the famous psychotherapist Carl Rogers, the computer asks 
  149. probing questions that make the patient open up, but it never 
  150. gives advice: the patient must discover the truth for herself.
  151.   I'll explain exactly how the program works, so you can become a 
  152. psychotherapist yourself, and get hundreds of thousands of 
  153. dollars by bilking your patients.
  154.                                          The computer begins by 
  155. replacing some of the patient's words:
  156. Typed by patient                                   Replacement
  157.                                            mom     mother
  158.                                            dad     father
  159.                                            dont    don't
  160.                                            cant    can't
  161.                                            wont    won't
  162.                                            dreamed dreamt
  163.                                            dreams  dream
  164.                                            I       you@
  165.                                            me      you
  166.                                            you     I
  167.                                            my      your
  168.                                            your    my
  169.                                            myself  yourself
  170.                                            yourselfmyself
  171.                                            I'm     you're
  172.                                            you're  I'm
  173.                                            am      are@
  174.                                            were    was
  175. For example, the sentence Well, my boyfriend made me come here 
  176. becomes Well, your boyfriend made you come here.
  177.                                          The computer hunts 
  178. through the revised sentence or paragraph for one of these 
  179. keywords. . . . 
  180. Category 8:                                  computer, computers, 
  181. machine, machines
  182. Category 7:                                  name
  183. Category 6:                                  alike, like, same
  184. Category 5:                                  remember
  185. Category 4:                                  dreamt
  186. Category 3:                                  dream, if
  187. Category 2:                                  everybody, everyone, 
  188. nobody, was, your
  189. Category 1:                                  always
  190. Category 0:                                  are, are@, because, 
  191.                                              can, certainly, 
  192.                                              deutsch, espanol, 
  193.                                              francais, hello, 
  194.                                              how, I, I'm, 
  195.                                              italiano, maybe, my, 
  196.                                              no, perhaps, sorry, 
  197.                                              what, when, why, 
  198.                                              yes, you@, you're
  199. If the computer finds several of those keywords, it chooses the 
  200. one in the highest category; if they lie in the same category, it 
  201. chooses the one the patient typed first.
  202.   Then it looks up the keyword in this phrasebook:
  203. Patient's keywordComputer's reply
  204. alike       In what way?
  205. always      Can you think of a specific example?
  206. are
  207.   are I . . .Why are you interested in whether I am . . . or not?
  208.   are . . . Did you think they might not be . . . ?
  209. are@
  210.   are you . . .Do you believe you are . . . ?
  211.   are       Why do you say ``am''?
  212. because     Is that the real reason?
  213. can
  214.   can I . . .You believe I can . . . , don't you?
  215.   can you . . .Whether or not you can . . . depends on you more 
  216. than on me.
  217. certainly   You seem quite positive.
  218. computer    Do computers worry you?
  219. computers   Do computers worry you?
  220. deutsch     I am sorry, I speak only English.
  221. dream       What does that dream suggest to you?
  222. dreamt . . .Really, . . . ?
  223. espanol     I am sorry, I speak only English.
  224. everybody   Really, everybody?
  225. everyone    Really, everyone?
  226. hello       How do you do ___ please state your problem.
  227. how         Why do you ask?
  228. I
  229.   I remind you ofIn what way?
  230.   I are . . .What makes you think I am . . . ?
  231.   I . . . youWhy do you think I . . . you?
  232.   I         We were discussing you ___ not me.
  233. if . . .    Do you think it's likely that . . . ?
  234. I'm . . .   What makes you think I am . . . ?
  235. italiano    I am sorry, I speak only English.
  236. like
  237.   (am,is,are,was)...likeIn what way?
  238.   like      (get a different keyword)
  239. machine     Do computers worry you?
  240. machines    Do computers worry you?
  241. maybe       You don't seem quite certain.
  242. my . . .    Why are you concerned over my . . . ?
  243. name        I am not interested in names.
  244. no          Are you saying ``no'' just to be negative?
  245. nobody      Really, nobody?
  246. perhaps     You don't seem quite certain.
  247. remember
  248.   you remember . . .Do you often think of . . . ?
  249.   do I remember . . .Did you think I would forget . . . ?
  250.   remember  (get a different keyword)
  251. same        In what way?
  252. sorry       Please don't apologize.
  253. was
  254.   was you . . .What if you were . . . ?
  255.   you was   Were you really?
  256.   was I . . .Would you like to believe I was . . . ?
  257.   was       (get a different keyword)
  258. what        Why do you ask?
  259. when        Why do you ask?
  260. why
  261.   why don't I . . .Do you believe I don't . . . ?
  262.   why can't you . . .Do you think you should be able to . . . ?
  263. yes         You seem quite positive.
  264. you@
  265.   you (want, need) . . .What would it mean to you if you got . . 
  266. . ?
  267.   you are . . . (sad, unhappy, depressed, sick)I am sorry to hear 
  268. you are (sad, etc.).
  269.   you are . . . (happy, elated, glad, better)How have I helped 
  270. you to be (happy, etc.)?
  271.   you (feel, think, believe, wish) youDo you really think so?
  272.   you (feel, think, believe, wish) . . . I(use the keyword ``I'' 
  273. instead)
  274.   you are . . .       Is it because you are . . . that you came 
  275. to me?
  276.   you (can't, cannot) . . .How do you know you can't . . . ?
  277.   you don't . . .     Don't you really . . . ?
  278.   you feel            Tell me more about such feelings.
  279.   you . . . I         Perhaps in your fantasy we . . . each 
  280. other.
  281. your
  282.   your . . . (mother, father, sister, brother, wife, 
  283. children)Tell me more about your family.
  284.   your . . .                Your . . . 
  285. you're                      (treat as ``you@ are'')
  286.   For example, if the keyword is sorry, the computer looks up 
  287. sorry in the phrasebook, which says to print ``Please don't 
  288. apologize.''
  289.                                                      Suppose the 
  290. patient types, ``If the job is lousy, he'll die.'' The keyword is 
  291. if. In the phrasebook, if is followed by three dots, which stand 
  292. for the part of the clause that comes after if, which is ``the 
  293. job is lousy''. (The computer figures out where the clause ends 
  294. by looking at the punctuation.) The phrasebook says to print ``Do 
  295. you think it's likely that the job is lousy?''
  296.                                                      The symbol @ 
  297. serves just to locate the correct keyword in the phrasebook. 
  298. Thereafter, it's ignored.
  299.                                                      Here's what 
  300. happens if the keyword is you@. After locating you@ in the 
  301. phrasebook, the computer ignores the @. If the patient's revised 
  302. sentence contains you want . . . or you need . . . , the computer 
  303. prints ``What would it mean to you if you got . . . ?'' If the 
  304. patient's sentence contains you are . . . sad, the computer 
  305. prints ``I am sorry to hear you are sad.''
  306.                                                      For each 
  307. reply in the phrasebook, Weizenbaum stored a list of 
  308. alternatives. For example, here are the alternatives to ``Please 
  309. don't apologize'':
  310. Apologies are not necessary.
  311. What feelings do you have when you apologize?
  312. I've told you that apologies are not required.
  313. While chatting with the patient, the computer keeps track of 
  314. which replies it has printed already, and uses the alternatives 
  315. to avoid repetition.
  316.                                                      If the 
  317. patient's statement doesn't contain a keyword, the computer may 
  318. give one of these replies:
  319. I am not sure I understand you fully.
  320. Please go on.
  321. What does that suggest to you?
  322. Do you feel strongly about discussing such things?
  323. Or it may take a second look at earlier parts of the 
  324. conversation, retrieve a clause that contained your . . . , and 
  325. print one of these replies:
  326. Let's discuss further why your . . . .
  327.  
  328. Earlier you said your . . . .
  329.  
  330. Does that have anything to do with the fact that your . . . ?
  331. For example, it may retrieve the clause Your boyfriend made you 
  332. come here and print:
  333. Does that have anything to do with the fact that your boyfriend 
  334. made you come here?
  335.                                                      The program 
  336. was written at the Massachusetts Institute of Technology (MIT). 
  337. The programmer, Joseph Weizenbaum, named the program ``Eliza'', 
  338. to honor Eliza Doolittle ___ the woman in My Fair Lady and 
  339. Pygmalion, who learned to speak English and have a good 
  340. conversation.
  341.   Some people think Weizenbaum's program shows that computers can 
  342. communicate as well as psychotherapists. But Weizenbaum himself 
  343. holds the opposite view; he believes the program shows that 
  344. psychotherapists communicate as poorly as computers.
  345.   You can buy a disk called Golden Oldies, which contains 
  346. imitations of four classic programs: Adventure, Pong, Life, and 
  347. Weizenbaum's Eliza. The disk is published by Software Country, 
  348. distributed by Electronic Arts, costs about $35, and comes in 
  349. versions for the Apple 2e & 2c, Macintosh, Commodore 64, Amiga, 
  350. Atari 800 XL, and IBM PC.
  351.   Improved versions After Weizenbaum invented Eliza, other 
  352. programmers invented improvements. For example, in one of the 
  353. improved versions, whenever the computer doesn't understand you, 
  354. it reacts like the typical rich psychotherapist ___ it bluffs and 
  355. says:
  356. I understand fully. Please go on.
  357.   The computer never really understands what you say. For 
  358. example, in one version, if you say ``My, my, my!'', the computer 
  359. will reply:
  360. Tell me more about your your your.
  361.   Radio Shack's headquarters phoned me, and asked for permission 
  362. to use my explanation of Weizenbaum's program. I agreed, 
  363. whereupon Radio Shack invented and sold a version that makes the 
  364. computer talk ___ by using a voice synthesizer. You'll freak out 
  365. when you hear the computer's voice giving you therapy. Radio 
  366. Shack's program costs $15, comes on tape, and requires a model 1 
  367. or model 3 TRS-80. To hear the voice, you must buy the voice 
  368. synthesizer. (But even if you don't buy the voice synthesizer, 
  369. the program will impress you.)
  370.                              The cheapest way to get a version of 
  371. Eliza is to buy a book called BASIC Computer Games Volume 2 
  372. (TRS-80 edition). Get it at your local Radio Shack store. It 
  373. costs just $7 and includes 84 programs ___ one of which is a 
  374. version of Eliza written in BASIC. That version of Eliza isn't as 
  375. good as the original; but since it's cheap and written in BASIC, 
  376. you'll enjoy buying it for self-study. Here's part of a 
  377. conversation between that version and a patient:
  378. Patient:                         I am not the greatest 
  379. conversationalist.
  380. Computer:                        Do you believe it is normal to 
  381.                                  be not the greatest 
  382.                                  conversationalist?
  383.  
  384. Patient:                         Yes.
  385. Computer:                        I see.
  386.  
  387. Patient:                         How can you see without any 
  388. eyes?
  389. Computer:                        Don't you believe that I can see 
  390. without any eyes?
  391.  
  392. Patient:                         No. So prove it.
  393. Computer:                        Are you saying ``no'' just to be 
  394. negative?
  395.  
  396. Patient:                         Absolutely not.
  397. Computer:                        You are being a bit negative.
  398.  
  399. Patient:                         So you want to fight, do you?
  400. Computer:                        We were discussing you ___ not 
  401. me.
  402.  
  403. Patient:                         Don't back down, friend.
  404. Computer:                        Why do you bring up the topic of 
  405. friends?
  406.  
  407. Patient:                         I don't have any friends.
  408. Computer:                        Don't you really have any 
  409. friends?
  410.  
  411. Patient:                         No, not really.
  412. Computer:                        Why not?
  413.                              Unfortunately, most versions of 
  414. Eliza are passive. The typical Eliza program works well only if 
  415. the patient's talkative. If the patient doesn't say much, the 
  416. program can't react too much and is useless.
  417.                              Recently, John Holland invented a 
  418. better version of Eliza. His version's more aggressive. He calls 
  419. it ``Analiza 2'' (because it ``analizas'' the patient). If the 
  420. patient's too quiet or too boring or too uninformative, Analiza 2 
  421. makes the computer give the patient a ``jolt'', by asking an 
  422. electrifying question or making a ridiculous Zen-like statement. 
  423. For example, the computer might ask, ``Did you ever try to commit 
  424. suicide with a hammer?'' Here's a list of the computer's favorite 
  425. ``jolts'':
  426. Nasty questions the computer asks
  427. Did you ever try to commit suicide with a hammer?
  428. Is your interest in sex still not very strong?
  429. Do you want to know what I suspect?
  430.  
  431. Prying questions the computer asks
  432. Tell me, what is your worst fear?
  433. Why don't you tell me about your first romance?
  434. I want to know about your lifestyle.
  435. Tell me what is troubling you.
  436. Honestly now, what is your biggest concern?
  437.  
  438. Ridiculous statements the computer makes
  439. With strong drugs you can learn to control your concern!
  440. You certainly are curious!
  441. And you are obstinate!
  442.  
  443. Complaints the computer makes
  444. You seem inhibited.
  445. This isn't really very exciting.
  446. Getting started is often difficult.
  447.  
  448. Comments the computer makes about timing and complexity
  449. Well, time is almost up. Is there anything else you want to 
  450. discuss?
  451. I hate to bring this up, but you still owe me $53.00 from your 
  452. last visit!
  453. Last time you were extremely inquisitive. Have you abandoned 
  454. seeking simple solutions?
  455. Last time you were difficult to understand. Were you trying to 
  456. confuse me?
  457. Let's keep it simple this time.
  458. Say, I am having trouble with your long sentences. Keep it 
  459. simple. I am only human.
  460.                              Yes, your patient will love to hear 
  461. the computer say it's ``only human''. You can get Analiza 2 for 
  462. $50 from Supersoft (Box 1628, Champaign, IL 61820, 217-359-2112). 
  463. It runs on most computers that have the CP/M operating system.
  464.   Paranoid What if the world becomes full of computerized 
  465. psychoanalysts, and there aren't enough patients to keep them 
  466. busy? Kenneth Colby solved that problem: in 1971 at Stanford 
  467. University, he programmed a computer to be a crazy patient. 
  468. Specifically, he programmed his computer to be a paranoid 
  469. schizophrenic. In the following conversations, psychiatrists 
  470. interrogate his computer.
  471. Conversation 1
  472. Psychiatrist:How long have you been in the hospital?
  473. Computer:About a week.
  474.  
  475. Psychiatrist:Are you depressed?
  476. Computer:No.
  477.  
  478. Psychiatrist:Were you taking drugs?
  479. Computer:No.
  480.  
  481. Psychiatrist:What is your problem?
  482. Computer:People make me nervous.
  483.  
  484. Psychiatrist:How do they do that?
  485. Computer:They stare at me.
  486.  
  487. Psychiatrist:Why?
  488. Computer:Maybe they don't like the way I look or something.
  489.  
  490. Psychiatrist:Do you look unusual in some way?
  491. Computer:Why are you interested in my looks?
  492.  
  493. Psychiatrist:Because you said people stare at you.
  494. Computer:Well, sometimes.
  495.  
  496. Psychiatrist:How do you get along with people?
  497. Computer:I don't communicate very much with the others.
  498.  
  499. Psychiatrist:Why not?
  500. Computer:They don't realize what's going on most of the time.
  501.  
  502. Psychiatrist:What is going on?
  503. Computer:Do you know how the underworld operates?
  504.  
  505. Conversation 2
  506. Psychiatrist:How are you today?
  507. Computer:OK.
  508.  
  509. Psychiatrist:What is your name?
  510. Computer:My name is Frank Smith.
  511.  
  512. Psychiatrist:How old are you?
  513. Computer:Twenty-eight.
  514.  
  515. Psychiatrist:Are you married?
  516. Computer:No.
  517.  
  518. Psychiatrist:What sort of work do you do?
  519. Computer:I work in the post office.
  520.  
  521. Psychiatrist:Why are you in the hospital?
  522. Computer:I shouldn't be here.
  523.  
  524. Psychiatrist:Who brought you here?
  525. Computer:The police.
  526.  
  527. Psychiatrist:What trouble did you have with the police?
  528. Computer:Cops don't do their job.
  529.  
  530. Psychiatrist:What do they do instead?
  531. Computer:Cops arrest the wrong people.
  532.  
  533. Psychiatrist:They arrest the right ones sometimes.
  534. Computer:The law should get those Italian crooks.
  535.  
  536. Psychiatrist:Are you Italian?
  537. Computer:My background is British but I was born in this country.
  538.  
  539. Psychiatrist:Do you know any Italian crooks?
  540. Computer:I try to avoid the underworld.
  541.  
  542. Psychiatrist:Are you afraid of the underworld?
  543. Computer:The Mafia are the worst gangsters.
  544.  
  545. Psychiatrist:Are you connected with the Mafia?
  546. Computer:You know, they know me.
  547.  
  548. Psychiatrist:How do they know you?
  549. Computer:They spy on me.
  550.  
  551. Psychiatrist:Why do they?
  552. Computer:They want to get false evidence to frame me.
  553. Psychiatrist:                                  What do they have 
  554. against you?
  555. Computer:                                      They are out to 
  556. get me.
  557.  
  558. Psychiatrist:                                  What for?
  559. Computer:                                      I have 
  560. incriminating evidence against the Mafia.
  561.  
  562. Psychiatrist:                                  Are you sure of 
  563. all this?
  564. Computer:                                      You don't believe 
  565. me?
  566.  
  567. Conversation 3
  568. Psychiatrist:                                  Tell me something 
  569. about yourself.
  570. Computer:                                      I went to the 
  571. track at Bay Meadows a while back.
  572.  
  573. Psychiatrist:                                  What happened 
  574. there?
  575. Computer:                                      There's a lot of 
  576. crooked gambling there.
  577.  
  578. Psychiatrist:                                  Why do you think 
  579. the gambling is crooked?
  580. Computer:                                      I have worried 
  581.                                                that gangsters 
  582.                                                have too much 
  583.                                                power in this 
  584.                                                country.
  585.  
  586. Conversation 4
  587. Psychiatrist:                                  How do you like 
  588. the hospital?
  589. Computer:                                      I don't 
  590. communicate very much with others.
  591.  
  592. Psychiatrist:                                  Do other people at 
  593. the hospital get on your nerves?
  594. Computer:                                      I shouldn't have 
  595. come here.
  596.  
  597. Psychiatrist:                                  Why do you say you 
  598. shouldn't have come here?
  599. Computer:                                      They made me come 
  600. here.
  601.                                          The computer's response 
  602. depends on how much FEAR, ANGER, and MISTRUST it has at the 
  603. moment. Those three variables are affected by what the 
  604. psychiatrist says. For example, if the psychiatrist gives the 
  605. computer a compliment, the FEAR, ANGER, and MISTRUST usually 
  606. decrease. But if MISTRUST is already high, the computer 
  607. interprets the compliment as a sly attempt at pacification, and 
  608. its ANGER increases instead of decreases. ANGER can also be 
  609. increased by questions that humiliate (Tell me about your sex 
  610. life) or imply subjugation (Perhaps you should stay in the 
  611. hospital longer).
  612.                                          To prove his computer is 
  613. paranoid, Colby had 8 psychiatrists interview it and also 
  614. interview paranoid humans via teletypewriters. Transcripts of all 
  615. the interviews were sent to psychiatrists around the country, who 
  616. were asked to judge whether each interview was with a human or 
  617. with the computer. The psychiatrists were unable to tell the 
  618. difference: only 51% of their guesses were correct.
  619.                                          Some computerists got 
  620. the ``brainstorm'' of hooking Weizenbaum's computer to Colby's, 
  621. to see whether the computerized psychotherapist could cure the 
  622. computerized schizophrenic. The experiment was a disaster: both 
  623. computers were so passive that the discussion rapidly degenerated 
  624. into trivia.
  625.                                          But so do conversations 
  626. between humans!
  627.  
  628.             FALL IN LOVE
  629.   Can the computer help you fall in love? Here are some famous 
  630. attempts, in chronological order. (I've rounded all dates to the 
  631. nearest 5 years.)
  632.  
  633.            TV love (1960)
  634.   A computer appeared on national TV, to make people fall in 
  635. love.
  636.   Guys and gals in the audience answered questionnaires about 
  637. their personality and fed them into the computer. The computer 
  638. chose the guy and gal that were most compatible. That guy and gal 
  639. had their first blind date on national television.
  640.   Each week, that scenario was repeated: the computer chose 
  641. another couple from the audience.
  642.   Each lucky couple appeared on the show again several weeks 
  643. later so the audience could find out whether the couple was in 
  644. love.
  645.   One of the couples was unhappy: the gal didn't like the guy, 
  646. even though she wanted to like him. She volunteered to be 
  647. hypnotized. So, on national TV, a hypnotist made her fall in love 
  648. with her partner.
  649.   The computer was a huge Univac. Today, the same kind of 
  650. matching could be done with a microcomputer. Any volunteers?
  651.  
  652.    Computer-dating services (1965)
  653.   College students began relying on computers, to find dates. 
  654. Here's how the typical computer-dating service worked. . . .
  655.   You answered a long questionnaire ___ about 8 pages. The 
  656. questionnaire asked about your sex, age, height, weight, hair 
  657. color, race, religion, how often you drank and smoked, how 
  658. ``handsome'' or ``attractive'' you were (on a scale of 1 to 10), 
  659. how far you wanted to go on your first date, whether you wanted 
  660. to get married soon, and how many children you'd like. It also 
  661. asked many questions about your personality.
  662.   One of the questions was:
  663. Suppose you receive in the mail some spoons you didn't order. The 
  664. accompanying note says the spoons were sent by a charitable 
  665. organization, and begs you to either send a contribution or 
  666. return the spoons. You don't like the spoons. What will you do?
  667. 1. Keep the spoons without paying.
  668. 2. Return the spoons.
  669. 3. Pay for the spoons.
  670.   Another question was:
  671. A girl returned from her date after curfew. Her excuse was that 
  672. her boyfriend's car broke down. What's your reaction?
  673. Again, you had a multiple-choice answer. One of the choices was, 
  674. ``Ha!''
  675.   For each question, you had to say how you would answer it, and 
  676. how you'd want your date to answer it. That was tough. What if 
  677. you wanted your date to be stunningly beautiful but also humble? 
  678. What if you wanted to meet somebody who's ugly and insecure 
  679. enough to be desperate to have sex? Such issues were debated in 
  680. college dorms throughout the nation.
  681.                                          After completing the 
  682. questionnaire, you mailed it with about $10 to the 
  683. computer-dating service. Within two months, the service would 
  684. send you the names, addresses, and phone numbers of at least 5 
  685. people you could date. If your personality was very easy to 
  686. match, the service might send you more than 5 names; but even if 
  687. your personality was lousy, you'd get at least 5. Periodically 
  688. throughout the year, you'd also get updates that matched you with 
  689. people who enrolled after you.
  690.                                          The most popular 
  691. computer-dating service was Operation Match, started by students 
  692. at Harvard. Its main competitor was Contact, started by students 
  693. at M.I.T. Both services quickly became profitable and had 
  694. subscribers from all across the country.
  695.                                          One gal's personality 
  696. was so wonderful that the computer matched her with 110 guys! She 
  697. had to explain to her mom why 110 guys were always on the phone 
  698. ___ and she had to figure out how to say ``no'' to 109 of them.
  699.                                          One gal got matched to 
  700. her roommate's boyfriend. They didn't stay roommates long.
  701.                                          When I was a freshman, I 
  702. applied to both services, to make sure I'd meet ``the gal of my 
  703. dreams''. Contact sent me names of gals at prestigious schools 
  704. (such as Wellesley and Bennington), while Operation Match sent me 
  705. names of gals at schools such as the State University of New York 
  706. at Albany.
  707.                                          I thought I was the only 
  708. nut desperate enough to apply to both services, but I got a 
  709. surprise! When I saw the list of names from Contact and the list 
  710. from Operation Match, I noticed a gal who appeared on both lists! 
  711. Like me, she'd been desperate enough to apply to both services, 
  712. and both computers agreed she'd be a perfect match for me!
  713.                                          I had a date with her 
  714. but couldn't stand her.
  715.                                          When I'd answered the 
  716. questionnaire, I was a very bashful boy, so the computer matched 
  717. me to bashful girls. But by the time I received the computer 
  718. printout, I'd become wilder, and the girls the computer 
  719. recommended were no longer ``my type''.
  720.                                          Contact raised its price 
  721. to $15, then $20. But $20 was still cheap for what you were 
  722. getting.
  723.                                          Contact ran a newspaper 
  724. ad that seemed to be selling groceries. It said, ``Dates ___ 2¢ 
  725. per pound''. The ad then explained that one gal got enough dates 
  726. so that, when she totaled the weight of their bodies, she figured 
  727. they cost her 2¢ per pound.
  728.  
  729.                                             The Dartmouth dater (1965)
  730.                                          When Dartmouth College 
  731. was still all-male, a student there wrote a cruel program that 
  732. evaluated dates by asking lots of ``practical'' questions such 
  733. as:
  734. Is she pretty?
  735. How far away does she live?
  736. Does she have a car?
  737.                                          I put down that I was 
  738. dating a 14-year-old girl who was 7 feet tall and weighed 300 
  739. pounds but had a perfect personality. I gave her personality a 
  740. 10, and even said that she lived nearby and had a car.
  741.                                          In spite of her 
  742. excellent personality, the computer didn't like her. The computer 
  743. said:
  744. She must be pregnant. Where did you get that pig?
  745. Worst score yet produced by this computer!
  746.  
  747.          Video dating (1975)
  748.   During the 1970's, people wanted everything to be natural. They 
  749. wanted ``natural food'' and ``natural love''.
  750.   Since computerized love seemed unnatural, its popularity 
  751. declined. Operation Match and Contact went out of business.
  752.   They were replaced by video dating, in which a video-dating 
  753. service shows you videotapes of members of the opposite sex and 
  754. lets you contact the person whose videotape you like best. That 
  755. way, you never have a ``blind'' date: you see the person on 
  756. videotape before you make the date. The service also makes a 
  757. videotape of you!
  758.   The video-dating service tapes thousands of people. Since you 
  759. don't have enough time to look at thousands of tapes, the service 
  760. tells you to answer a questionnaire, which is fed into a 
  761. computer. The computer tells you which people you're most 
  762. compatible with; then you look at those people's tapes.
  763.  
  764.        Computer dancing (1975)
  765.   At a Connecticut prep school (Hotchkiss), the head of the 
  766. computer center arranged a ``computer dance''.
  767.   All the students answered questionnaires, which were fed into a 
  768. computer. The computer matched the boys with the girls, so each 
  769. boy got one girl. The boy had to take the girl to the dance.
  770.   The computer center's staff announced the dancing partners in a 
  771. strange way: one morning, the students found all the halls 
  772. decorated with strips of punched paper tape, saying (in 
  773. billboard-style letters) messages such as ``George Smith & Mary 
  774. Jones''. If you were a student, you looked up and down the halls 
  775. (your heart beating quickly), to find the tape displaying your 
  776. name alongside the name of your mysterious computer lover.
  777.   Shrieks and groans. ``Aarrgghh! You wouldn't believe who the 
  778. computer stuck me with!''
  779.  
  780.       Computer weddings (1980)
  781.   Here's how the first true ``computer marriage'' occurred. . . . 
  782.   One company's terminal was attached to another company's 
  783. computer. A programmer at the first company often asked a 
  784. programmer at the second company for help. They contacted each 
  785. other by typing messages on their terminals, and let the computer 
  786. relay the messages back and forth. One of the programmers was a 
  787. guy, the other was a gal, and they fell in love, even though they 
  788. had never met. Finally, the guy typed on his terminal, ``Let's 
  789. get married''. The gal typed back, ``Yes''. And so they got 
  790. engaged ___ even though they had never met.
  791.   Their marriage ceremony used three terminals: one for the guy, 
  792. one for the gal, and one for the minister. The minister typed the 
  793. questions at his own terminal; then the guy and gal typed back, 
  794. ``I do''.
  795.   Reverend Apple Reverend Apple is an Apple computer programmed 
  796. to perform marriage ceremonies.
  797.   It performed its first marriage on Valentine's Day, 1981. The 
  798. groom was a guy named Richard; the bride was a gal named Debbie. 
  799. The computer printed the standard wedding-ritual text on the 
  800. screen, and then asked the usual questions. Instead of answering 
  801. ``I do'', the bride and groom just had to type ``Y''.
  802.   Reverend Apple is smart. For example, if the bride or groom 
  803. types ``N'' instead of ``Y'', the computer beeps, tells the 
  804. couple to try again, and repeats the question.
  805.                                          The program was written 
  806. by M.E. Cavanaugh at the request of Rev. Jon Jaenisch, who stood 
  807. by Reverend Apple while the ceremony was being performed.
  808.                                          Rev. Jaenisch is a 
  809. minister of the Universal Life Church ___ the church that lets 
  810. you become an ``ordained minister'' by just paying $5, and become 
  811. a ``doctor of divinity'' by just paying $20. He's known as the 
  812. ``Archbishop in Charge of Keyboarding''.
  813.                                          For his next feat, he 
  814. plans to make the computer perform divorces. He also uses the 
  815. computer to persuade kids to come to church. He claims, ``What 
  816. better way to get kids into church than by letting them play with 
  817. a computer? It's more interesting than praying.''
  818.                                          For a while, he couldn't 
  819. interest enough couples in using Reverend Apple. He complained, 
  820. ``It's not easy to convince people to get married by a computer. 
  821. They don't think it's romantic.'' NBC television news and many 
  822. newspapers wanted to interview him, but he couldn't find enough 
  823. willing couples.
  824.                                          And besides, he's a 
  825. reverend only part-time. His main job's as an employment agent: 
  826. he's supposed to help companies find programmers. He thought 
  827. Reverend Apple's reputation would help him find programmers, but 
  828. it didn't.
  829.                                          But Reverend Apple 
  830. eventually started to catch on. During its first eight months, it 
  831. performed six marriages.
  832.                                          Jaenisch says, ``The 
  833. first couple had nothing to do with computers professionally: the 
  834. groom drove a tow-truck and was an hour late for the ceremony 
  835. because he wanted to work overtime. But the second couple was 
  836. very involved with computers: they even asked for a printout of 
  837. the ceremony.''
  838.                                          The sixth ceremony's 
  839. groom earned his living by fixing computer power supplies and 
  840. said, ``It was nice with our friends all gathered around the 
  841. console, and someone brought champagne. But part of our vow was 
  842. to never buy a home computer: we have to get away from machines 
  843. sometime.''
  844.  
  845.                                                   Love Bug (1980)
  846.                                          You can buy a Love Bug. 
  847. It's a small computerized box that you put in your pocket. You 
  848. feed the box information about your personality. When you walk 
  849. through a singles bar, if you get near a person of the opposite 
  850. sex who's compatible and has a Love Bug also, your Love Bug 
  851. beeps. As you and the other person get closer and closer, the 
  852. Love Bugs beep to each other even more violently. The more 
  853. violently your Love Bug beeps, the closer you are to your ideal 
  854. partner.
  855.                                          Using a Love Bug to find 
  856. a date is like using a Geiger counter to find uranium. The louder 
  857. the Love Bug beeps, the louder your heart will pound.
  858.  
  859.                                                Selectrocution (1980)
  860.                                          If you don't like the 
  861. Love Bug, how about a love billboard? One company sells love 
  862. billboards to singles bars.
  863.                                          Each person who enters 
  864. the bar wears a gigantic name tag showing the person's initials. 
  865. For example, since I'm Russ Walter, my tag says, in gigantic 
  866. letters, ``RW''. If I see an attractive gal whose tag says 
  867. ``JN'', and I like her smile, I tell the person who operates the 
  868. billboard. A few seconds later, a gigantic computerized billboard 
  869. hanging over the entire crowd flashes this message:
  870. FOR JN FEMALE:  YOU HAVE A NICE SMILE--RW MALE
  871. Everybody in the bar sees my message. When the gal of my dreams, 
  872. ``JN female'', sees it, she hunts for ``RW male'', and we unite 
  873. in computerized joy.
  874.                                          That's great for bashful 
  875. people, like me, who'd rather pass notes than face a stranger 
  876. unprepared.
  877.   It's called Selectrocution, because it gives your social life 
  878. an electronic tingle that ends all your problems.
  879.  
  880.           Interlude (1980)
  881.   The most provocative sex program is Interlude. It interviews 
  882. both you and your lover, then tells you what sexual activities to 
  883. perform. Some of the activities are quite risqué. (Puritans think 
  884. the program should be called ``Inter Lewd''.)
  885.   The program runs on your Radio Shack or Apple computer. (The 
  886. explicit full-color ad shows a half-clad girl on satin sheets 
  887. caressing her Apple.)
  888.   The program's based loosely on Masters-and-Johnson sexual 
  889. therapy. It interviews each person separately and privately, then 
  890. recommends a sexual interlude.
  891.   During the interview, the computer asks you questions such as:
  892. How long would you like the interlude to last?
  893. You can choose any length of time, from ``several seconds'' to 
  894. ``several days''.
  895.   If you choose ``several seconds'', the computer recommends that 
  896. while driving home from a party, you put your lover's finger in 
  897. your mouth and seductively caress it with your tongue. If you 
  898. choose ``several days'', the computer recommends telling your 
  899. lover to meet somebody at the airport; but when your lover 
  900. arrives at the airport, make your lover find you there instead, 
  901. armed with two tickets for a surprise vacation.
  902.   The computer also asks questions such as:
  903. Do you like surprises?
  904. You have several choices: you like to give surprises, be 
  905. surprised, or don't like surprises at all. If you like to be 
  906. surprised, and your lover likes to give surprises, the computer 
  907. tells you to leave the room; after you've left, the computer 
  908. gives your lover secret hints about the best way to surprise you.
  909.   The computer asks which parts of the body you like. (One choice 
  910. is: ``buttocks''.) The computer also asks which kinds of 
  911. accessories you like. (One choice is: ``whips and chains''.) The 
  912. computer asks whether you want the interlude to occur 
  913. ``immediately'' or ``later'': if you say ``later'', the computer 
  914. recommends buying elaborate props to make the interlude fancier.
  915.   Some of the interludes are weird. For example, if you're a 
  916. woman and want to surprise your husband, the computer recommends 
  917. calling his office to invite him home for lunch. When he arrives, 
  918. he finds all the shades pulled down: you do a nude dance on the 
  919. table, then sit down to eat.
  920.   During the interview, the computer's questions are often corny. 
  921. For example, the computer asks:
  922. If your interlude were on TV, what show would it resemble?
  923. Sample choices are ``Three's Company'', ``Roots'', and ``a 
  924. commercial''. If you say ``Roots'', the computer says ``heavy!'' 
  925. If you say ``a commercial'', the computer says ``yecch!''
  926.   The computer asks how much sex you'd like. If you say ``lots!'' 
  927. but your lover says the opposite, the computer will recommend 
  928. that you take a cold shower, to cool your hot passion.
  929.                                          If you've been married 
  930. for at least twenty years, you'd probably like to change a few 
  931. things about your sex life but are afraid to tell your spouse 
  932. that you've been less than thrilled. You'd like an intermediary 
  933. to whom you can express your anxieties and who will pass the 
  934. message to your spouse gently. The Interlude program acts as that 
  935. intermediary, in a playful way.
  936.                                          Interlude's programmer 
  937. says he created it because he was tired of hearing people wonder 
  938. what to do with their personal computers. Once you've tried the 
  939. Interlude program, your personal computer will suddenly become 
  940. very personal!
  941.                                          It's rated R. To avoid 
  942. an X rating, it insists on having one man and one woman: it 
  943. doesn't permit homosexuality, group sex, or masturbation. Sorry!
  944.                                          The program came out in 
  945. May, 1980. Within a year, ten thousand copies were sold.
  946.                                          In 1986, an improved 
  947. version was invented: Interlude 2. It's available for the IBM PC 
  948. and the Apple 2 family. You can get it for $45.95 (plus $4.95 
  949. shipping and $1.78 for credit-card processing) from Dolphin 
  950. Computers (309 Judah Street #214, San Francisco, CA 94122, phone 
  951. 415-566-4400).
  952.  
  953.                                                  Pornopoly (1980)
  954.                                          To have an orgy, try 
  955. this trick. Invite your friends over for a ``game''. Tell them 
  956. it's a computerized version of Monopoly. When they arrive, 
  957. surprise them by telling them they'll play Pornopoly, the 
  958. computerized version of Monopoly that's rated X.
  959.                                          Like Monopoly, Pornopoly 
  960. lets you buy and sell property; but the streets have names such 
  961. as Bedroom Avenue, Horny Avenue, Hot Jugs Avenue, Jock Strap 
  962. Place, and Orgasm Railroad. You get penalty cards such as: name 7 
  963. four-letter words that rhyme with duck. You might be told to play 
  964. doctor, and conduct a physical examination of another player . . 
  965. . or remove the pants of your favorite player by using only your 
  966. teeth. When a player lands on a monopoly that you own, the player 
  967. must take a drink, remove an article of clothing, kiss you, give 
  968. you a free feel, or strip completely for two turns. At the end of 
  969. the game, whoever remains dressed is the winner.
  970.                                          This successful program 
  971. has been featured on national TV. Copies have been requested by 
  972. Hugh Hefner, Johnny Carson, Rona Barrett, an army chaplain, and a 
  973. dozen foreign countries. To add your own name to that list, try 
  974. contacting Computer Consultants of Iowa (Box 427, Marion, Iowa 
  975. 52302, 319-373-1306, if still in business).
  976.                                          Pornopoly costs $30, but 
  977. the company doesn't accept money: it accepts only Master Charge, 
  978. Visa, and COD. If you're a kid, tough luck: the company says, 
  979. ``This is an adult party game rated XXX and some people may find 
  980. it offensive.''
  981.                                          Among the offended is a 
  982. New Orleans grandmother who read an article about the program and 
  983. wrote this note to the company: ``Thanks to you, I intend to 
  984. start contributing to Moral Majority, something I've avoided 
  985. until now.''
  986.                                          The program's available 
  987. for Radio Shack, Apple, Commodore, and Atari computers. Infoworld 
  988. (the microcomputer industry's scandal sheet) criticizes the Atari 
  989. version for its poor graphics, vague manual, and occasional bugs. 
  990. If you try the Radio Shack, Apple, or Commodore version, tell me 
  991. how you like it. And can I play?
  992.  
  993.            REPLACE PEOPLE
  994.   Computers can replace people.
  995.  
  996.              Bartenders
  997.   Many bar owners don't trust the bartenders they hire. They 
  998. claim the bartenders give too many free drinks to friends, steal 
  999. money from the till, and put too much or too little liquor in the 
  1000. drinks.
  1001.   To solve the problem, many bars now contain a computer that 
  1002. mixes and pours drinks. The computer mixes accurately. Although 
  1003. the computer is run by the bartender, the computer keeps an 
  1004. accurate record of how many drinks it makes, so there is little 
  1005. chance for cheating. The computer also keeps track of the 
  1006. inventory.
  1007.   The computers are manufactured and sold by NCR (Dayton, Ohio), 
  1008. Bar Boy Inc. (San Diego, California), Electronic Dispensers 
  1009. International (Concord, California), and Anker-Werke (Germany). 
  1010. Prices range from $600 to $15000. Holiday Inn has been developing 
  1011. its own model.
  1012.  
  1013.                Doctors
  1014.   If you're ill, would a computer diagnose your illness more 
  1015. accurately than a human doctor?
  1016.   During the 1970's this article appeared in The Times:
  1017.   A medical diagnostic system designed at Leeds University has 
  1018. proved more accurate than doctors in assessing the most likely 
  1019. cause of acute abdominal pain among patients admitted to the 
  1020. university's department of surgery.
  1021.   Last year 304 such patients were admitted to the unit, and the 
  1022. computer's diagnosis proved correct in 92 percent of the cases, 
  1023. compared with 80 percent accuracy by the most senior doctor to 
  1024. see each case.
  1025.   After each patient had been seen by the doctor and examined, 
  1026. the doctor's findings were passed on to a technician, who 
  1027. translated them into language used by the computer. The computer 
  1028. would list the likely diagnoses in order of probability. If the 
  1029. computer and the doctor in charge of the case disagreed, the 
  1030. computer would on request suggest further investigations that 
  1031. might be useful.
  1032.   In the year-long trial the computer's diagnoses proved correct 
  1033. in 279 cases. In 15 it was wrong, in 8 the patient's condition 
  1034. was not included in the diseases considered by the computer, and 
  1035. in 2 no computer diagnosis was made because the doctors concerned 
  1036. with the case disagreed about the findings.
  1037.   Whereas the computer advised an operation on 6 occasions when 
  1038. it would have proved unnecessary, in practice 30 such operations 
  1039. were carried out on the basis of the surgeon's own judgment. The 
  1040. computer accurately classified 84 of the 85 patients with 
  1041. appendicitis, compared with 75 by the doctors, and its suggestion 
  1042. that no operation was necessary proved correct on 136 out of 137 
  1043. occasions.
  1044.   The computer is reliable only if accurate data are fed into it 
  1045. on the basis of the doctor's interrogation and examination of the 
  1046. patient.
  1047.   In a 1970's experiment at another hospital, each patient was 
  1048. given a computer terminal and could use it to communicate with 
  1049. the computer about his illness. Afterwards, most of the patients 
  1050. said they'd rather deal with the computer than with the human 
  1051. doctors.
  1052.   This report is by James Martin and Adrian Norman:
  1053.   The Los Angeles County General Hospital Shock Research Unit is 
  1054. concerned with circulatory shock, which results in low blood 
  1055. pressure, low blood flow, and an extremely unstable circulatory 
  1056. system. The proportion of circulatory-shock patients who die is 
  1057. large.
  1058.   A computer was installed to improve the monitoring and care of 
  1059. these patients. A patient in this ward has many instruments 
  1060. attached to him: six thermistors measure the temperature at 
  1061. various parts of his body; pressures are measured in the 
  1062. circulatory system; catheters are inserted into an artery and a 
  1063. vein; and a urinometer measures liquid waste. The instruments are 
  1064. all directly connected to the computer, and other details about 
  1065. the patient are given to the machine by the nurse.
  1066.                                          The computer prints out 
  1067. a log about the patient's condition, and if any measurement falls 
  1068. outside critical limits set by the doctor for that patient, the 
  1069. computer notifies the ward staff.
  1070.  
  1071.                                                       Tutors
  1072.                                          Judith Harris, John 
  1073. Swets, and Wallace Feurzeig made the computer run a logic game. 
  1074. The computer thinks of a letter of the alphabet; your goal is to 
  1075. discover which letter it is, as quickly as possible. You can ask 
  1076. the computer questions about the letter's shape.
  1077.                                          In this example, the 
  1078. human begins by asking whether the letter has any curves:
  1079. Human:                                       Curves?
  1080. Computer:                                    One.
  1081.  
  1082. Human:                                       Loose ends?
  1083. Computer:                                    Two.
  1084.  
  1085. Human:                                       Obliques?
  1086. Computer:                                    Zero.
  1087.  
  1088. Human:                                       C.
  1089. Computer:                                    You don't have 
  1090.                                              enough information 
  1091.                                              yet to get the right 
  1092.                                              answer. How do you 
  1093.                                              know it isn't J, for 
  1094.                                              example?
  1095.  
  1096. Human:                                       J.
  1097. Computer:                                    You don't have 
  1098.                                              enough information 
  1099.                                              yet. How do you know 
  1100.                                              it isn't C, for 
  1101.                                              example?
  1102.  
  1103. Human:                                       Horizontals?
  1104. Computer:                                    Zero. That was a 
  1105.                                              wasted question. It 
  1106.                                              did not eliminate 
  1107.                                              any possibilities.
  1108.  
  1109. Human:                                       Verticals?
  1110. Computer:                                    Zero. Good question.
  1111.  
  1112. Human:                                       G.
  1113. Computer:                                    Wrong. Your guess is 
  1114.                                              inconsistent with 
  1115.                                              the information 
  1116.                                              you've been given. 
  1117.                                              You must have 
  1118.                                              forgotten that the 
  1119.                                              right letter hasn't 
  1120.                                              any HORIZONTAL 
  1121.                                              LINES. The letter 
  1122.                                              you guessed can't be 
  1123.                                              right ___ it has too 
  1124.                                              many LOOSE ENDS.
  1125.  
  1126. Human:                                       Curves?
  1127. Computer:                                    The answer is still: 
  1128. one.
  1129.  
  1130. Human:                                       X.
  1131. Computer:                                    Wrong. Your guess is 
  1132.                                              inconsistent with 
  1133.                                              the information 
  1134.                                              you've been given. I 
  1135.                                              think you only made 
  1136.                                              that guess to see 
  1137.                                              what I would say.
  1138.  
  1139. Human:                                       C.
  1140. Computer:                                    Yes, you've done it. 
  1141.                                              Good work. If you 
  1142.                                              want to play again, 
  1143.                                              start right in.
  1144.                                          Notice that the human 
  1145. can input either a guess (followed by a period) or a question 
  1146. (followed by a question mark). Possible guesses are A, B, C, D, 
  1147. etc. Possible questions are Curves, Loose ends, etc.; the human 
  1148. is given a list of possible questions before he runs the program.
  1149.                                          Medicine That logic game 
  1150. was modified to form a lesson about medicine. The medical case 
  1151. was designed by Preston Munter, of the Harvard Health Center; the 
  1152. programming was done by John Swets, Wallace Feurzeig, and Myra 
  1153. Breen. The object of the ``game'' is to diagnose an illness 
  1154. correctly, by performing sufficient tests. The human can input 
  1155. either a diagnosis (Measles) or a request for a test (X-ray). 
  1156. Before running the program, the human is given a list of 35 
  1157. possible diagnoses and 40 possible tests.
  1158.   Here are excerpts from a long RUN:
  1159. Computer:Please type your name, the date, and a slash.
  1160.  
  1161. Human:Myra Breen, August 1963/
  1162. Computer:It is mid-December. Late this afternoon, a patient was 
  1163.       admitted to the hospital. He is a well-developed, 
  1164.       well-nourished, 34-year-old Negro coal miner. He was 
  1165.       brought in by his wife, who reports that he has had "the 
  1166.       grippe" for about 4 days. During the past 3 days she has 
  1167.       given him fluids and aspirin every 6 hours with no sign of 
  1168.       improvement. Today he suddenly became much worse. For the 
  1169.       past 2 days he had complained of feeling very warm and of 
  1170.       pain in his belly. This morning he had a severe shaking 
  1171.       chill that was so bad his teeth rattled. . . . Begin the 
  1172.       physical examination as soon as you are ready.
  1173.  
  1174. Human:Physical exam?
  1175. Computer:Fine. What in particular would you like to know?
  1176.  
  1177. Human:Chest exam?
  1178. Computer:It is common practice in the medical profession to begin 
  1179.       an examination by checking the patient's general 
  1180.       appearance.
  1181.  
  1182. Human:General appearance?
  1183. Computer:Acutely prostrated, lying in bed on his right side. 
  1184.       Looks acutely ill, sweaty, apprehensive, but alert. 
  1185.       Paroxysmal cough, bringing up a little rusty streaked 
  1186.       sputum. Breathing with difficulty.
  1187.  
  1188. Human:EENT exam?
  1189. Computer:Check all of the vital signs first!
  1190.  
  1191. Human:Temperature?
  1192. Computer:103.
  1193.  
  1194. Human:Blood pressure?
  1195. Computer:150/84.
  1196.  
  1197. Human:Pulse?
  1198. Computer:100 a minute. Occasional extra systole. . . . 
  1199.  
  1200. Human:Abdominal exam?
  1201. Computer:Diffusely tender, some splinting noticed in lower right 
  1202.       quadrant and occasionally radiating to right flank area.
  1203.  
  1204. Human:The grippe.
  1205. Computer:There's insufficient information for any conclusive 
  1206.       diagnosis at this time. You're not being careful!
  1207.  
  1208. Human:Musculo-skeletal system?
  1209. Computer:Essentially negative, except for generalized weakness 
  1210.       and aching.
  1211.  
  1212. Human:Extremities?
  1213. Computer:Normal, except nail beds are questionably cyanotic but 
  1214.       difficult to determine because of patient's skin color.
  1215.  
  1216. Human:Rectal exam?
  1217. Computer:You're not being careful! The patient is too acutely ill 
  1218.       for a rectal examination.
  1219.  
  1220. Human:Laboratory reports?
  1221. Computer:Fine. Which lab tests would you like to have done?
  1222.  
  1223. Human:RBC?
  1224. Computer:A red blood count will be done. Results will be sent to 
  1225.       you as soon as possible. . . . Would you like to alert the 
  1226.       night nurse to set up the lab tests you'll want to see 
  1227.       tomorrow?
  1228.  
  1229. Human:Yes.
  1230. Computer:Good. That will speed things up. However, there's 
  1231.       nothing further that can be done tonight to help the 
  1232.       patient. Let's call it quits for now. Perhaps you'll be in 
  1233.       a better position to diagnose the case tomorrow after the 
  1234.       first lab reports come in and you have reexamined the 
  1235.       patient. Have a good night's sleep, and when you're back on 
  1236.       duty tell me to ``proceed.''
  1237.  
  1238. Human:Proceed.
  1239. Computer:It's morning and you're making your rounds. You've come 
  1240.       to our patient and are given the night nurses's report: 
  1241.       ``Temperature remained high but patient spent a reasonably 
  1242.       comfortable night. Cough diminished but persists. 
  1243.       Occasional traces of blood in sputum. Patient still 
  1244.       dyspneic and still has pain in abdomen. The second series 
  1245.       of lab tests was initiated.'' Some of the first series of 
  1246.       lab reports have just come in: urine exam, insufficient 
  1247.       fluid for analysis; RBC, 5 million; WBC, 15500; 
  1248.       differential, shift to left. . . . You may be prepared to 
  1249.       make the diagnosis now. Are you?
  1250. Human:                                       No.
  1251. Computer:                                    Fine. Nevertheless, 
  1252.                                              we've got a fair 
  1253.                                              amount of 
  1254.                                              information ___ at 
  1255.                                              least enough to 
  1256.                                              raise some, perhaps 
  1257.                                              several, diagnostic 
  1258.                                              possibilities. Have 
  1259.                                              you any in mind?
  1260.  
  1261. Human:                                       Yes.
  1262. Computer:                                    Name one.
  1263.  
  1264. Human:                                       Pulmonary 
  1265. infarction.
  1266. Computer:                                    Fine. Pleural pain, 
  1267.                                              dyspnea, fever, and 
  1268.                                              the physical exam 
  1269.                                              signs and symptoms 
  1270.                                              are certainly 
  1271.                                              consistent with 
  1272.                                              pulmonary 
  1273.                                              infarction. Do you 
  1274.                                              think that shaking 
  1275.                                              chills and the 
  1276.                                              presence of rusty 
  1277.                                              sputum further 
  1278.                                              support this 
  1279.                                              diagnosis?
  1280.  
  1281. Human:                                       No.
  1282. Computer:                                    Right. Though the 
  1283. white count is consistent, isn't it?
  1284.  
  1285. Human:                                       No.
  1286. Computer:                                    Wrong. The white 
  1287.                                              count is consistent. 
  1288.                                              This should further 
  1289.                                              support your 
  1290.                                              diagnosis. We've 
  1291.                                              considered one 
  1292.                                              possibility. Do you 
  1293.                                              have another 
  1294.                                              diagnosis in mind?
  1295.  
  1296. Human:                                       No.
  1297. Computer:                                    In that case I'd 
  1298.                                              like to talk about 
  1299.                                              viral pneumonia. The 
  1300.                                              tachycardia, high 
  1301.                                              WBC, elevated 
  1302.                                              respiratory rate, 
  1303.                                              shaking chills, 
  1304.                                              bloody sputum, and 
  1305.                                              severe pleural pain 
  1306.                                              all lend weight to 
  1307.                                              that diagnosis ___ 
  1308.                                              right?
  1309.  
  1310. Human:                                       Yes.
  1311. Computer:                                    You're way off base. 
  1312.                                              There's not an iota 
  1313.                                              of truth to any part 
  1314.                                              of what I just said; 
  1315.                                              in fact, the 
  1316.                                              opposite of what I 
  1317.                                              said is true. You'd 
  1318.                                              better review some 
  1319.                                              elementary medicine, 
  1320.                                              doctor. . . . 
  1321.  
  1322. Human:                                       Pleural effusion.
  1323. Computer:                                    A good contender, 
  1324.                                              but let's be 
  1325.                                              careful. What 
  1326.                                              evidence would rule 
  1327.                                              it in or out?
  1328.  
  1329. Human:                                       Positive evidence of 
  1330. another disease.
  1331. Computer:                                    Is there such 
  1332. evidence?
  1333.  
  1334. Human:                                       Possible evidence 
  1335. though inconclusive.
  1336. Computer:                                    Since you aren't 
  1337.                                              sure, continue the 
  1338.                                              examination until 
  1339.                                              you have conclusive 
  1340.                                              evidence one way or 
  1341.                                              the other.
  1342.  
  1343. Human:                                       RBC?
  1344. Computer:                                    An RBC is not 
  1345.                                              indicated at this 
  1346.                                              time. It's an 
  1347.                                              unnecessary expense 
  1348.                                              for a poor patient.
  1349.  
  1350. Human:                                       WBC?
  1351. Computer:                                    22500.
  1352.  
  1353. Human:                                       Staph pneumonia.
  1354. Computer:                                    Fair enough. But 
  1355.                                              where is the guilty 
  1356.                                              organism? When you 
  1357.                                              have isolated 
  1358.                                              staphylococcus 
  1359.                                              aureus in 
  1360.                                              significant numbers 
  1361.                                              in your samples of 
  1362.                                              this patient's 
  1363.                                              fluids or 
  1364.                                              excretions, let me 
  1365.                                              know.
  1366.  
  1367. Human:                                       Pneumococcal 
  1368. pneumonia.
  1369. Computer:                                    Are you absolutely 
  1370. sure?
  1371.  
  1372. Human:                                       Yes.
  1373. Computer:                                    You can't be, until 
  1374.                                              you have irrefutable 
  1375.                                              evidence. What would 
  1376.                                              help you nail down 
  1377.                                              the diagnosis?
  1378.  
  1379. Human:                                       Evidence from sputum 
  1380. smear and culture.
  1381. Computer:                                    Good. Anything else?
  1382.  
  1383. Human:                                       No.
  1384. Computer:                                    I think you might 
  1385.                                              also want evidence 
  1386.                                              from x-ray findings. 
  1387.                                              Go ahead and we'll 
  1388.                                              see what happens.
  1389.  
  1390. Human:                                       X-ray?
  1391. Computer:                                    Not ready yet. Try 
  1392. the blood lab.
  1393.  
  1394. Human:                                       Differential?
  1395. Computer:                                    Marked shift to 
  1396. left. . . . 
  1397.  
  1398. Human:                                       X-ray?
  1399. Computer:                                    The x-ray shows: 
  1400.                                              area of 
  1401.                                              consolidation, right 
  1402.                                              lower lobe, 
  1403.                                              consistent with 
  1404.                                              pneumatic process. 
  1405.                                              This is a good time 
  1406.                                              to review the total 
  1407.                                              findings and to make 
  1408.                                              another tentative 
  1409.                                              diagnosis ___ if you 
  1410.                                              like.
  1411.  
  1412. Human:                                       Evidence of fluid?
  1413. Computer:                                    No evidence of 
  1414. fluid.
  1415.  
  1416. Human:                                       Sputum smear?
  1417. Computer:                                    Gram-positive rods.
  1418.  
  1419. Human:                                       Strep pneumonia.
  1420. Computer:                                    Fine, but how can 
  1421.                                              you distinguish this 
  1422.                                              from any other 
  1423.                                              bacterial pneumonia? 
  1424.                                              Wait until you have 
  1425.                                              an irrefutable case.
  1426. Human:Viral pneumonia.
  1427. Computer:But what about shaking chills, rusty sputum, elevated 
  1428.       WBC, etc. ___ all the evidence of an acute febrile 
  1429.       bacterial process? You are in error, doctor.
  1430.  
  1431. Human:Sputum culture?
  1432. Computer:Presence of pneumococci.
  1433.  
  1434. Human:Strep pneumonia.
  1435. Computer:Incredible! You have disregarded the positive findings 
  1436.       that should clearly have established the correct diagnosis.
  1437.  
  1438. Human:Pneumococcal pneumonia.
  1439. Computer:Very good. It's obvious now, though there were some 
  1440.       confusing signs and symptoms in the picture. You've seen 
  1441.       your way through to the correct diagnosis.
  1442.   Religion Bruce Jones programmed the computer to teach you about 
  1443. religion. Here's an example:
  1444. Computer:This program is designed to teach you some things about 
  1445.       the so-called "documentary hypothesis" as formulated by 
  1446.       Julius Wellhausen and others. First, what is your name?
  1447.  
  1448. Human:George
  1449. Computer:All right, George, who wrote the first five books of the 
  1450.       Bible, according to tradition?
  1451.  
  1452. Human:Moses
  1453. Computer:You are off to a good start. Do you know what those 
  1454.       first five books are called? They are called . . . 
  1455.  
  1456. Human:the Pentateuch
  1457. Computer:Right you are. Torah is the traditional Jewish name for 
  1458.       them. They are also called the Pentateuch ("penta" = 5). 
  1459.       People very early noticed that there was a problem with 
  1460.       saying that Moses wrote the Torah. If you already know the 
  1461.       basis of their suspicion, tell me.
  1462.  
  1463. Human:Some things in the text reflect later conditions, after 
  1464. Moses.
  1465. Computer:You are right, of course, but there was an even more 
  1466.       obvious problem. Try again.
  1467.  
  1468. Human:I don't know.
  1469. Computer:At the end of the Pentateuch the author says, ``Moses . 
  1470.       . . died there.'' Does that seem strange to you?
  1471.  
  1472. Human:yes
  1473. Computer:Why?
  1474.  
  1475. Human:Moses could not have written about his own death.
  1476. Computer:O.k. Also, right after the report of Moses' death and 
  1477.       burial (Deut. 34), the text says, ``No man knows the place 
  1478.       of his burial TO THIS DAY.'' How do those last three words 
  1479.       strike you?
  1480.  
  1481. Human:They seem later than the event they are talking about.
  1482. Computer:A good observation, George. Many people considered these 
  1483.       words to be anachronistic. That is, ``this day'' is sharply 
  1484.       distinguished from the time of the burial. It seems likely, 
  1485.       therefore, that the time of writing was much later than the 
  1486.       time of burial. Again (so the argument goes), Moses was not 
  1487.       the author. Would these observations about the end of 
  1488.       Deuteronomy convince you that Moses did not write the 
  1489.       Torah?
  1490.  
  1491. Human:yes
  1492. Computer:Why do you say that?
  1493.  
  1494. Human:A man cannot write his own obituary.
  1495. Computer:Probably not many people would have been convinced if 
  1496.       our evidence were so limited. After all, Moses could have 
  1497.       written everything except the last chapter. However, as 
  1498.       early as the 18th century, people noticed another problem. 
  1499.       . . . 
  1500. The computer searched through the human's input, to see whether 
  1501. he used words indicating a correct answer. The computer never 
  1502. said a flat ``Wrong'', since religious answers are a matter of 
  1503. personal belief, and since the human might be smarter or weirder 
  1504. than the computer program was prepared for.
  1505.                                                       Robots
  1506.                                          In 1962 at MIT, Heinrich 
  1507. Ernst connected the computer to a mechanical hand that could 
  1508. feel. He made the hand build objects out of blocks, and made it 
  1509. put blocks into boxes.
  1510.                                          Shakey One of the most 
  1511. famous robots is a guy named ``Shakey'', built at the Stanford 
  1512. Research Institute (SRI) in 1970. His eye contains a television 
  1513. camera (optical scanner). Instead of legs, he has wheels. Instead 
  1514. of arms, he has antennae (for feeling) and a bumper (for 
  1515. pushing). His brain is a computer: instead of carrying it around 
  1516. with him, he leaves it in another room and communicates with it 
  1517. by wireless methods.
  1518.                                          To see how he works, 
  1519. suppose you type this message on his computer's terminal:
  1520. Push the block off the platform.
  1521. He begins by looking for the platform. If the platform is not in 
  1522. the room, he goes out to the hall and steers himself through the 
  1523. hall (by looking at the baseboards) until he arrives at the next 
  1524. room. He peers into the room to see whether it contains a 
  1525. platform. If not, he hunts for another room. When he finally 
  1526. finds a room containing a platform with a block on it, he tries 
  1527. to climb onto the platform to push the block off. But before 
  1528. climbing the platform, he checks the platform's height. If it's 
  1529. too high to get onto easily, he looks for a device to help him 
  1530. climb it. For example, if a ramp is lying in the room, he pushes 
  1531. the ramp next to the platform and then wheels himself up the 
  1532. ramp. Finally, he pushes the block off.
  1533.                                          He can handle unexpected 
  1534. situations. For example, while he's getting the ramp, suppose you 
  1535. pull the platform to a different place. That doesn't faze him: he 
  1536. hunts for the platform again, and then pushes the ramp to it.
  1537.                                          In 1971, Shakey's powers 
  1538. were extended, so he can handle commands such as:
  1539. Turn on the lightswitch.
  1540. If the lightswitch is too high for his bumper to reach, he looks 
  1541. for a device to climb onto, such as a box. If he finds a box that 
  1542. looks helpful, he climbs onto it to check whether it is tall 
  1543. enough; if it is, he climbs off, pushes it to the lightswitch, 
  1544. climbs on it again, and finally flicks the switch.
  1545.                                          Another task he can 
  1546. handle is:
  1547. Push three boxes together.
  1548. He finds the first box and pushes it to the second. Then he finds 
  1549. the third box, and pushes it to the second.
  1550.                                          He understands over 100 
  1551. words. Whatever command you give him becomes his ``goal'', and he 
  1552. must reason out how to accomplish it. He might discover that to 
  1553. accomplish it, he must accomplish another goal first ___ for 
  1554. example, to move the block off the platform, he must first find 
  1555. the platform; to do that, he might have to look in another room; 
  1556. to do that, he must leave the room he's in; to do that, he must 
  1557. turn his wheels.
  1558.   Simulator One Here's a picture of a robot named Simulator One:
  1559.   In the picture, a doctor is taking Simulator One's blood 
  1560. pressure and pulse. Another doctor is watching the computer 
  1561. console.
  1562.   Simulator One is a model patient. He can blink, breathe, cough, 
  1563. vomit, respond to drugs, and even die. He's used in med school, 
  1564. to train doctors how to administer anesthetics during surgery.
  1565.   Improved robots This report (abridged) is by Bertram Raphael, 
  1566. the director of the SRI Artificial Intelligence Center:
  1567.   Here's what robots were capable of doing a few years ago.
  1568.   At Hitachi Central Research Laboratory, a TV camera was aimed 
  1569. at an engineering plan drawing of a structure. A second camera 
  1570. looked at blocks spread out on a table. The computer 
  1571. ``understood'' the drawing, reached toward the blocks with its 
  1572. arm, and built the structure.
  1573.   At MIT, the camera was not shown a plan; instead, it was shown 
  1574. an example of the actual structure desired. The computer figured 
  1575. out how the structure could be constructed, and then built an 
  1576. exact copy.
  1577.   At Stanford University, the hand obeyed spoken directions. For 
  1578. example, if someone said into the microphone, ``Pick up the small 
  1579. block on the left,'' that is precisely what the arm would do.
  1580.   In Scotland at the University of Edinburgh, a jumble of parts 
  1581. for two wooden toys was placed on a table. ``Freddy,'' the 
  1582. Edinburgh robot, spread out the parts so that it could see each 
  1583. one clearly, and then, with the help of a vise-like work station 
  1584. at one corner of the table, assembled first the toy car and then 
  1585. the toy boat.
  1586.   Recently, robot researchers have built robots that can perform 
  1587. truly practical tasks. For example:
  1588.   At Stanford, the system that used to stack toy blocks can now 
  1589. assemble a real water pump.
  1590.   At SRI, a computer-controlled arm with touch and force sensors 
  1591. can feel its way as it packs assembled pumps into a case.
  1592.   At MIT, programs are under development to enable a computer to 
  1593. inspect and repair circuit boards for use in computers, TV sets, 
  1594. and other electronic equipment.
  1595.   The Beast Not all robots involve computers. Here's an example 
  1596. of a noncomputerized robot (reported by James Slagle, abridged):
  1597.   A. George Carlton, John G. Chubbuck, and others at the Applied 
  1598. Physics Laboratory of John Hopkins University built a machine 
  1599. called The Beast.
  1600.   It's a battery-operated cylinder on wheels that's 18 inches in 
  1601. diameter. It has tactile, sonar, and optical apparatus. The sonar 
  1602. permits The Beast to find its way down the center of the hall. 
  1603. When its battery becomes sufficiently run-down, The Beast 
  1604. optically looks for an electric outlet and plugs itself in to 
  1605. recharge its battery.
  1606.   The Beast was often let loose to roam in the halls and offices 
  1607. at the Applied Physics Laboratory in order to see how long it 
  1608. could survive without ``starving.'' Once it survived 40.6 hr. 
  1609. Many a new and unsuspecting secretary has been startled when The 
  1610. Beast entered her office, plugged itself into an electric outlet, 
  1611. and then departed.
  1612.   When it feels a step down, it knows enough to turn around, so 
  1613. that it doesn't fall downstairs. But this logic sometimes makes 
  1614. it starve when it encounters a raised threshold. After getting on 
  1615. the threshold, it thinks it's about to fall, so it turns around. 
  1616. After turning around it again thinks it's going to fall, so it 
  1617. turns back and forth until it starves.
  1618.   It also starved when some workmen changed all the outlets from 
  1619. the flush to the projecting type. To cope with the new situation, 
  1620. the researchers changed some of the circuitry.
  1621.  
  1622.                                                      Japan A 
  1623. newspaper article said that in Japan robots are being used in 
  1624. many practical ways. One robot arc-welds, reducing the time by 
  1625. 90%. Another grasps an object, determines the best way to pack it 
  1626. in a box, and does the packing; it uses television cameras and 
  1627. delicate arms. Another washes windows. Another wiggles a rod to 
  1628. catch a fish, takes the fish off the hook, dumps it into a bin, 
  1629. and returns the line to the water. Another directs traffic. 
  1630. Talking robots are being used instead of kimono-clad females in 
  1631. inns and restaurants.
  1632.                                                      Commenting 
  1633. on the quality of life in Japan, the article went on to say that 
  1634. people are buying whiffs of oxygen from vending machines.
  1635.                                                      The article 
  1636. was tacked on the bulletin board at the MIT Artificial 
  1637. Intelligence Laboratory, together with this graffito about how 
  1638. the Japanese robots would act differently if they were as smart 
  1639. as people. . . . 
  1640. Human:                                                   Weld 
  1641. these parts.
  1642. Robot:                                                   The 
  1643.                                                          steel in 
  1644.                                                          those 
  1645.                                                          parts is 
  1646.                                                          grossly 
  1647.                                                          inferior
  1648.                                                          . They 
  1649.                                                          must 
  1650.                                                          have 
  1651.                                                          been 
  1652.                                                          made in 
  1653.                                                          the U.S. 
  1654.                                                          Send 
  1655.                                                          them 
  1656.                                                          back. 
  1657.                                                          Also, 
  1658.                                                          have my 
  1659.                                                          welding 
  1660.                                                          machine 
  1661.                                                          tuned 
  1662.                                                          up.
  1663.  
  1664. Human:                                                   Pack 
  1665. those widgets.
  1666. Robot:                                                   Can I 
  1667. break them in half first?
  1668.  
  1669. Human:                                                   No.
  1670. Robot:                                                   ``No'' 
  1671.                                                          is not 
  1672.                                                          an 
  1673.                                                          allowed 
  1674.                                                          answer. 
  1675.                                                          I only 
  1676.                                                          have 
  1677.                                                          small 
  1678.                                                          shipping 
  1679.                                                          boxes. 
  1680.                                                          (Proceed
  1681.                                                          s to 
  1682.                                                          break 
  1683.                                                          widgets 
  1684.                                                          in half 
  1685.                                                          and 
  1686.                                                          stuff 
  1687.                                                          them 
  1688.                                                          into 
  1689.                                                          boxes.)
  1690.  
  1691. Human:                                                   Wash 
  1692. those windows.
  1693. Robot:                                                   What?? 
  1694.                                                          And get 
  1695.                                                          my 
  1696.                                                          precious 
  1697.                                                          electric
  1698.                                                          al and 
  1699.                                                          mechanic
  1700.                                                          al 
  1701.                                                          componen
  1702.                                                          ts all 
  1703.                                                          wet??
  1704.  
  1705. Human:                                                   Catch a 
  1706. fish.
  1707. Robot:                                                   
  1708.                                                          (Proceed
  1709.                                                          s to 
  1710.                                                          catch a 
  1711.                                                          fish, 
  1712.                                                          take it 
  1713.                                                          off the 
  1714.                                                          hook, 
  1715.                                                          and 
  1716.                                                          throw it 
  1717.                                                          back.) 
  1718.                                                          Okay.
  1719.  
  1720. Human:                                                   What did 
  1721. you throw it back for?
  1722. Robot:                                                   It was 
  1723.                                                          under 
  1724.                                                          the size 
  1725.                                                          limit. 
  1726.                                                          Anyway, 
  1727.                                                          it was 
  1728.                                                          full of 
  1729.                                                          mercury.
  1730.  
  1731. Human:                                                   Direct 
  1732. that traffic.
  1733. Robot:                                                   Will I 
  1734.                                                          get 
  1735.                                                          hazardou
  1736.                                                          s duty 
  1737.                                                          pay for 
  1738.                                                          this 
  1739.                                                          work?
  1740.  
  1741. Human:                                                   Greet 
  1742. those customers.
  1743. Robot:                                                   Sorry, 
  1744. my shift is over.
  1745.  
  1746. Human:                                                   (Ready 
  1747. to faint.) Give me some air.
  1748. Robot:                                                   Use that 
  1749.                                                          "machine
  1750.                                                          " over 
  1751.                                                          there. 
  1752.                                                          (Points 
  1753.                                                          to 
  1754.                                                          vending 
  1755.                                                          machine.
  1756.                                                          )
  1757.  
  1758.  
  1759.               BE POETIC
  1760.   Programmers have taught the computer to compose its own poetry 
  1761. and other literature.
  1762.  
  1763. The Taoist Hiroshima Elevator (1961)
  1764.   In Italy, Nanni Balestrini created poetry by feeding the 
  1765. computer three passages.
  1766.   He fed the computer this passage from Lao Tzu's Tao Te Ching: 
  1767. ``While the multitude of things comes into being, I envisage 
  1768. their return. Although things flourish, they all return to their 
  1769. roots.''
  1770.   He fed this passage from Michihito Hachiya's Hiroshima Diary: 
  1771. ``The blinding fireball expands rapidly, thirty times brighter 
  1772. than the sun. When it reaches the stratosphere, the summit of the 
  1773. cloud takes on the well-known mushroom shape.''
  1774.   He fed this from Paul Goldwin's The Mystery of the Elevator: 
  1775. ``head pressed on shoulder, hair between lips, lay motionless 
  1776. without speaking, till he moved his fingers slowly, trying to 
  1777. grasp''.
  1778.   He told the computer to mix them up. Here's the result:
  1779. Head pressed on shoulder, thirty times
  1780. brighter than the sun, I envisage their return ___ 
  1781. until he moved his fingers slowly. . . . And while the multitude
  1782. of things comes into being, at the summit of the cloud
  1783. they all return to their roots and take on
  1784. the well-known mushroom shape, endeavoring to grasp.
  1785.  
  1786. Hair between lips, they all return
  1787. to their roots. In the blinding fireball
  1788. I envisage their return ___ until he moved his fingers
  1789. slowly ___ and although things flourish
  1790. takes on the well-known mushroom shape, endeavoring
  1791. to grasp, while the multitude of things comes into being.
  1792.  
  1793. In the blinding fireball I envisage
  1794. their return. When it reaches the stratosphere, while the 
  1795. multitude
  1796. of things comes into being, head pressed
  1797. on shoulder, thirty times brighter than the sun,
  1798. they all return to their roots. Hair
  1799. between lips takes on the well-known mushroom shape.
  1800.  
  1801. They lay motionless without speaking. Thirty times
  1802. brighter than the sun, they all return
  1803. to their roots. Head pressed on shoulder,
  1804. they take on the well-known mushroom shape, endeavoring
  1805. to grasp. And although things flourish,
  1806. they expand rapidly, hair between lips.
  1807.  
  1808. While the multitude of things comes into being in the blinding
  1809. fireball, they all return
  1810. to their roots. They expand rapidly ___ until he moved
  1811. his fingers slowly, when it reached the stratosphere
  1812. and lay motionless without speaking, thirty times
  1813. brighter than the sun, endeavoring to grasp.
  1814.  
  1815. I envisage their return ___ until he moved his fingers
  1816. slowly in the blinding fireball.
  1817. They all return to their roots. Hair
  1818. between lips and thirty times brighter than the sun
  1819. lay motionless without speaking. They expand,
  1820. rapidly endeavoring to grasp the summit.
  1821.   In Balestrini's program, Tao Te Ching was represented as four 
  1822. phrases:
  1823. while the multitude of things comes into being
  1824. I envisage their return
  1825. although things flourish
  1826. they all return to their roots
  1827. Hiroshima Diary was represented as six phrases, and The Mystery 
  1828. of the Elevator as five.
  1829.                                          For each verse, the 
  1830. computer was told to choose nine phrases at random, and print 
  1831. them in a random order (never juxtaposing phrases from the same 
  1832. passage), to form six lines of roughly equal metrical length.
  1833.                                          Actually the computer 
  1834. printed the poem in capital letters, without punctuation; 
  1835. Balestrini himself then added the punctuation and polished the 
  1836. grammar. The whole thing was actually done in Italian; you've 
  1837. been reading Edwin Morgan's translation, with my punctuation.
  1838.  
  1839.                                                   Bubbles (1966)
  1840.                                          At Northwestern 
  1841. University, programmers made the computer compose nice poetry. To 
  1842. use their program, you type a list of nouns, verbs, and other 
  1843. words. The computer randomly chooses five of your words to be 
  1844. theme words. The computer combines all your words to form 
  1845. sentences, but chooses the theme words more often than the 
  1846. others. It combines the sentences into verses and tries to keep 
  1847. the lengths of the lines approximately equal. It puts a theme 
  1848. word into the title.
  1849.                                          In one poem, the 
  1850. computer chose bubble to be a theme word. The title was: ODE TO A 
  1851. BUBBLE. The poem contained phrases such as, ``Ah, sweet bubble.'' 
  1852. The word bubble appeared so often that even the stupidest reader 
  1853. could say:
  1854.                                          ``Oh, yeah. I really 
  1855. understand this poem. Ya see, it's about a bubble.''
  1856.                                          The poem had all the 
  1857. familiar poetic trappings, such as ``but alas!'', which marked 
  1858. the turning point. (Cynics argue that the poem didn't really have 
  1859. a turning point, since the computer didn't have the faintest idea 
  1860. of what it was saying!)
  1861.        Kids and physics (1968)
  1862.   In England at Manchester University, Mendoza made the computer 
  1863. write children's stories. Here's a story the computer composed:
  1864.   The sun shone over the woods. Across the fields softly drifted 
  1865. the breeze, while then the clouds, which calmly floated all 
  1866. afternoon, moved across the fields.
  1867.   Squirrel, who scampered through the trees, quickly ran off; and 
  1868. off noisily ran Little Grey Rabbit. She sniffed at the house; but 
  1869. out of the door noisily hurried Hare, who peered at slowly the 
  1870. flowers. Squirrel quickly scampered over the woods and fields, 
  1871. but Old Grey Owl flew over the woods and fields. Down the path to 
  1872. the woods ran Little Grey Rabbit, who then sniffed at a 
  1873. strawberry pie.
  1874.   The first paragraph uses these words:
  1875. Nouns
  1876. the clouds              1 1 0 1 0 0 1 0
  1877. the sun                 0 1 1 1 1 0 1 1
  1878. the breeze              1 1 0 1 1 2 0 0
  1879. the sky                 0 0 0 0 1 0 1 1
  1880.  
  1881. Adverbs
  1882. gently                  1 1 1 1 1 1 1 1
  1883. quietly                 1 1 1 1 1 1 1 1
  1884. loudly                  1 1 1 1 1 1 1 1
  1885. softly                  1 1 1 1 1 1 1 1
  1886. calmly                  1 1 1 1 1 1 1 1
  1887. soon                    1 1 1 1 1 1 1 1
  1888. then                    1 1 1 1 1 1 1 1
  1889. (no adverb)             2 2 2 2 2 2 2 2
  1890.  
  1891. Endings
  1892. by                      1 1 0 1
  1893. over the woods          1 1 1 1
  1894. across the meadows      1 1 1 1
  1895. through the trees       1 1 1 1
  1896. down                    0 0 1 0
  1897. for a long time         0 0 1 1
  1898. all day                 1 1 1 1
  1899. all afternoon           1 1 1 1
  1900. the grass                       1 1 1 1
  1901. the leaves of the trees         1 1 1 1
  1902. the garden                      1 1 1 1
  1903. the flowers                     1 1 1 1
  1904. the little house                1 0 1 1
  1905. the old oak tree                1 1 1 1
  1906. the treetops                    1 1 1 1
  1907.  
  1908. ADDITIONAL WORDS: which, and, while, they, it
  1909.   To construct a sentence, the computer uses that table. Here's 
  1910. how.
  1911.   First, the computer randomly chooses a noun. Suppose it chooses 
  1912. the sun.
  1913.   Then it looks across the row marked the sun, to choose a verb 
  1914. whose score isn't 0. For example, it's possible that the sun 
  1915. shone, but not possible that the sun melted. Suppose it chooses 
  1916. shone.
  1917.   Then it looks down the column marked shone, to choose an adverb 
  1918. and an ending. Notice that the ending can't be by, since its 
  1919. score is 0. No adverb has a score of 2, whereas gently has a 
  1920. score of 1; that makes no adverb twice as likely as gently.
  1921.   If the computer chooses no adverb and over the woods, the 
  1922. resulting sentence is: The sun shone over the woods. In fact, 
  1923. that's the first sentence of the story you just read.
  1924.                                          The computer 
  1925. occasionally changes the word order. For example, instead of 
  1926. typing ``The breeze drifted softly across the fields'', the 
  1927. computer begins the second sentence by typing, ``Across the 
  1928. fields softly drifted the breeze''.
  1929.                                          To combine short 
  1930. sentences into long ones, the computer uses the words at the 
  1931. bottom of the table: which, and, while, they, and it. If two 
  1932. consecutive clauses have the same subject, the computer 
  1933. substitutes a pronoun: they replaces the clouds; it replaces the 
  1934. sun, the trees, and the sky. The program says a which clause can 
  1935. come after a noun (not a pronoun); the which clause must use a 
  1936. different verb than the main clause.
  1937.                                          Here's the vocabulary 
  1938. for the second paragraph:
  1939. Nouns
  1940. Little Grey Rabbit         0 0 2 3 1 1 0 0
  1941. Old Grey Owl               0 3 0 0 1 3 2 2
  1942. Squirrel                   3 0 1 1 1 1 3 3
  1943. Hare                       0 0 0 2 1 1 2 2
  1944.  
  1945. Adverbs
  1946. then                       0 1 1 1 1 1 0 0
  1947. slowly                     0 2 0 0 1 1 1 1
  1948. quickly                    1 1 1 1 0 0 1 1
  1949. soon                       1 0 1 1 0 0 1 1
  1950. happily                    1 0 0 1 0 0 1 1
  1951. gaily                      1 0 0 1 0 0 1 1
  1952. noisily                    1 0 1 1 0 0 2 3
  1953. (no adverb)                5 4 4 5 2 2 5 5
  1954.  
  1955. Endings
  1956. off                        1 1 1 1
  1957. over the woods and fields  1 1 1 1
  1958. through the trees          1 1 1 1
  1959. among the treetops         0 1 0 0
  1960. into the home              1 0 1 1
  1961. out of the door            1 0 1 1
  1962. down the path to the woods 1 0 1 1
  1963. about the garden           1 1 1 1
  1964. the house                          1 1 0 0
  1965. the hollow tree                    1 1 0 0
  1966. an old oak tree                    1 1 0 0
  1967. the flowers                        1 1 0 0
  1968. two buns                           1 1 1 1
  1969. a strawberry pie                   1 1 1 1
  1970. six cabbages                       1 1 1 1
  1971.  
  1972. ADDITIONAL WORDS: who, and, but, she, he
  1973.                                          Here's another story the 
  1974. program produced:
  1975.                                          The breeze drifted by. 
  1976. Across the fields softly moved the clouds; and then the breeze, 
  1977. which calmly touched the treetops, drifted across the fields. 
  1978. Quietly the sun shone over the woods. The sky calmly shone across 
  1979. the fields.
  1980.                                          Out of the door ran 
  1981. Squirrel; and off hurried Hare, who munched and crunched two buns 
  1982. happily. Off slowly flew Old Grey Owl, and Squirrel soon ate two 
  1983. buns. Old Grey Owl, who peered at a strawberry pie, munched and 
  1984. crunched two buns; but noisily Little Grey Rabbit, who peered at 
  1985. an old oak tree, slowly ran down the path to the woods. Soon she 
  1986. hurried down the path to the woods, but then she sniffed at two 
  1987. buns. She hurried down the path to the woods.
  1988.  
  1989.   Why did Mendoza make the computer write those stories? He 
  1990. explains:
  1991.   This work all began when a well-known scientist joined our 
  1992. physics department. He had spent several years away from academic 
  1993. life and was able to take a long cool look at academic 
  1994. procedures. He soon formed the theory that students never learned 
  1995. any ideas; all they learned was a vocabulary of okay words which 
  1996. they strung together in arbitrary order, relying on the fact that 
  1997. an examiner pressed for time would not actually read what they 
  1998. had written but would scan down the pages looking for these 
  1999. words. I set out to test his hypothesis.
  2000.   I began by writing ``Little Grey Rabbit'' stories. I tested 
  2001. these stories out on my very small children; but after some 
  2002. minutes they grew irritable, because nothing actually happened. 
  2003. This shows that even small children of three can measure entropy.
  2004.   Then I altered the vocabulary and grammar ___ making the 
  2005. sentences all very dead ___ to imitate the style of physics 
  2006. textbooks. The endpoint came when a colleague at another 
  2007. university secretly sent me an exam a week before it was given to 
  2008. the students. I wrote vocabularies and copied down what the 
  2009. computer emitted. Using a false name, I slipped my paper in among 
  2010. the genuine ones. Unfortunately, it was marked by a very 
  2011. conscientious man, who eventually stormed into the Director's 
  2012. office shouting, ``Who the hell is this man ___ why did we ever 
  2013. admit him?'' So perhaps my colleague's hypothesis was wrong, and 
  2014. students are a little better than we think.
  2015.   Here's one of the computer's answers:
  2016. In electricity, the unit of resistance is defined by 
  2017. electrolysis; and the unit of charge, which was fixed at the 
  2018. Cavendish lab in Rayleigh's classic experiments, was measured at 
  2019. the Cavendish lab. Theoretically, the absolute ohm is defined in 
  2020. a self-consistent way. The unit of resistance, which was 
  2021. determined with a coil spinning in a field, was fixed at the 
  2022. Cavendish lab; and this, by definition, is expressed in 
  2023. conceptual experiments. Theoretically the absolute ohm, which was 
  2024. redetermined using combined e.m.u. and e.s.u., is expressed by 
  2025. the intensity at the center of a coil.
  2026.   Here's another of the computer's answers:
  2027.   In this country, Soddy considered Planck's hypothesis from a 
  2028. new angle. Einstein 50 years ago asserted quantisation.
  2029.   At a photocathode, electrons which undergo collisions in the 
  2030. Compton effect as energy packets or quanta are emitted at 
  2031. definite angles; nevertheless, particles in a photocell produce 
  2032. photoelectrons of energy hv=E0. Photons in vacuo transmute into 
  2033. lower frequencies, and light quanta in the Compton effect emit 
  2034. emission currents.
  2035.   Particles emit current proportional to energy; electrons in 
  2036. vacuo interact with loss of surface energy (work function); 
  2037. nevertheless, particles which are emitted in a photocell with 
  2038. conservation experimentally are conserved with energy hv. The 
  2039. former, at a metal surface, undergo collisions with emission of 
  2040. current; and at a metal surface, electrons produce emission 
  2041. currents.
  2042.   Einstein assumed the gas of quantum particles; but quite 
  2043. recently Rayleigh, who quite recently solved the problem in an 
  2044. old-fashioned way, considered radiation classically. Planck, who 
  2045. this century assumed the A and B coefficients, explained the gas 
  2046. of quantum particles but before Sommerfield; Rayleigh, who quite 
  2047. recently was puzzled on Boltzmann statistics, tackled the problem 
  2048. with disastrous results.
  2049.   Planck, who assumed the gas of quantum particles in 1905, this 
  2050. century considered the ultraviolet catastrophe; but quite 
  2051. recently Jeans, who tackled the problem in an old-fashioned way, 
  2052. was puzzled with disastrous results.
  2053.   Black body radiation that exerts thermodynamic forces in an 
  2054. engine is equivalent to a relativistic system. Out of a black 
  2055. body, a photon that is equivalent to (out of a black body) an 
  2056. assembly of photons is assumed to be a non-conservative system; 
  2057. at the same time, thermodynamically, black body radiation that in 
  2058. a piston is assumed to be a relativistic system exerts quantised 
  2059. forces.
  2060.   The radiation gas that obeys Wien's displacement law is 
  2061. considered as a system of energy levels. Quantally, a quantum 
  2062. particle exerts a Doppler-dependent pressure, although this 
  2063. produces equilibrium transition probabilities.
  2064.   Black body radiation in an engine produces equilibrium 
  2065. transition probabilities.
  2066.  
  2067.           Aerospace (1968)
  2068.   In 1968, Raymond Deffrey programmed the computer to write fake 
  2069. reports about the aerospace industry. Shortly afterwards, I 
  2070. improved the program. The improved program contains these lists:
  2071. Introductory phrases
  2072. thus
  2073. indeed
  2074. however
  2075. moreover
  2076. similarly
  2077. furthermore
  2078. for example
  2079. in addition
  2080. in particular
  2081. to some extent
  2082. in this regard
  2083. on the other hand
  2084. for the most part
  2085. as a resultant implication
  2086. in view of system operation
  2087. in respect to specific goals
  2088. based in system engineering concepts
  2089. utilizing the established hypotheses
  2090. based on integral subsystem considerations
  2091. considering the postulated interrelationships
  2092.  
  2093. Noun phrases
  2094. the structural design
  2095. the sophisticated hardware
  2096. the total system rationale
  2097. any discrete configuration made
  2098. the fully integrated test program
  2099. any associated supporting element
  2100. the product configuration baseline
  2101. the independent function principle
  2102. the preliminary qualification limit
  2103. the subsystem compatibility testing
  2104. the greater flight-worthiness concept
  2105. a constant flow of effective information
  2106. the characterization of specific criteria
  2107. the anticipated third-generation equipment
  2108. initiation of critical subsystem development
  2109. the evolution of specifications over a given time
  2110. the philosophy of commonality and standardization
  2111. the incorporation of additional mission constraints
  2112. a consideration of system and/or subsystem technologies
  2113. a large portion of the interface coordination communication
  2114.  
  2115. Verb phrases
  2116. adds explicit performance limits to
  2117. effects a significant implementation to
  2118. adds overriding performance constraints to
  2119. presents extremely interesting challenges to
  2120. is further compounded, when taking into account
  2121. must utilize and be functionally interwoven with
  2122. requires considerable systems analysis to arrive at
  2123. necessitates that urgent consideration be applied to
  2124. maximizes the probability of success and minimizes time for
  2125. recognizes the importance of other systems and necessity for
  2126.                                          To produce a typical 
  2127. sentence, the computer prints an introductory phrase, then a noun 
  2128. phrase, then a verb phrase, then a noun phrase. The phrases are 
  2129. chosen randomly.
  2130.                                          Each paragraph consists 
  2131. of six such sentences. The computer isn't allowed to use the same 
  2132. phrase twice within a paragraph. The introductory phrase is 
  2133. omitted from the first sentence of the first paragraph, the 
  2134. second sentence of the second paragraph, etc.; so the report 
  2135. can't begin with the word furthermore, and the style varies.
  2136.                                          Here's the beginning of 
  2137. one such report:
  2138.                                The Economic Considerations of the Aerospace Industry
  2139.  
  2140.                                          A large portion of the 
  2141. interface coordination communication necessitates that urgent 
  2142. consideration be applied to the product configuration baseline. 
  2143. For example, the fully integrated test program adds explicit 
  2144. performance limits to the independent function principle. 
  2145. Moreover, the sophisticated hardware presents extremely 
  2146. interesting challenges to the philosophy of commonality and 
  2147. standardization. In view of system operation, a constant flow of 
  2148. effective information must utilize and be functionally interwoven 
  2149. with the preliminary qualification limit. In addition, any 
  2150. discrete configuration made adds overriding performance 
  2151. constraints to any associated supporting element. Thus, the 
  2152. anticipated third-generation equipment maximizes the probability 
  2153. of success and minimizes time for the total system rationale.
  2154.  
  2155.      Me-Books (1972)
  2156.   In 1972, Freeman Gosden Jr. started the Me-Books Publishing 
  2157. Company. It published books for kids. But if you bought a Me-Book 
  2158. for your child, you wouldn't see in it the traditional names 
  2159. ``Dick, Jane, and Sally''; instead, you'd see the name of your 
  2160. own child. To order the book, you had to tell the company the 
  2161. names of all your children, and their friends, and pets. Their 
  2162. names appeared in the story.
  2163.   The story was printed beautifully, in a 32-page hard-covered 
  2164. book with pictures in color. It cost just $3.95.
  2165.   You could choose from four stories: ``My Friendly Giraffe'', 
  2166. ``My Jungle Holiday'', ``My Birthday Land Adventure'', and ``My 
  2167. Special Christmas''.
  2168.   For example, if you lived on Jottings Drive, and your 
  2169. daughter's name was Shea, and her friend's name was Douglas, the 
  2170. story ``My Friendly Giraffe'' included paragraphs such as this:
  2171.   One morning Shea was playing with Douglas in front of her home. 
  2172. When she looked up, what do you think she saw walking down the 
  2173. middle of Jottings Drive? You guessed it. A giraffe!
  2174.   Ted Nelson, author of Computer Lib, played a trick. He ordered 
  2175. a copy of ``My Friendly Giraffe'', but pretended that his child's 
  2176. name was ``Tricky Dick Nixon'' who lived on ``Pennsylvania Ave.'' 
  2177. in ``Washington''. Sure enough, the company sent him ``My 
  2178. Friendly Giraffe: A Me-Book for Tricky Dick''. Here are some 
  2179. excerpts:
  2180.   Once upon a time, in a place called Washington, there lived a 
  2181. little boy named Tricky Dick Nixon. Now, Tricky Dick wasn't just 
  2182. an ordinary little boy. He had adventures that other little boys 
  2183. and girls just dream of. This is the story of one of his 
  2184. adventures. It's the story of the day that Tricky Dick met a 
  2185. giraffe. . . . 
  2186.   As the giraffe came closer and closer, Tricky Dick started to 
  2187. wonder how in the world he was going to look him in the eye. . . 
  2188.   Tricky Dick knew there were no jungles in Washington. 
  2189. Especially on Pennsylvania Ave. But Tricky Dick wasn't even a 
  2190. little bit worried. First, because he was a very brave little 
  2191. boy. And second, because he knew that his friend, the giraffe, 
  2192. would never take him anyplace bad. . . . 
  2193.   Tricky Dick was home. Back in Washington. Back on Pennsylvania 
  2194. Ave. And with a story to tell his friends, that they wouldn't 
  2195. have believed if they hadn't seen Tricky Dick riding off on the 
  2196. giraffe's back. Tricky Dick would long be a hero to those who had 
  2197. seen him that day. . . . 
  2198.   There would be many other exciting adventures for Tricky Dick 
  2199. and his friends. And maybe, just maybe, if you're a very good 
  2200. boy, someday we'll tell you about those, too.
  2201.   Me-Books were commercially successful: many thousands of them 
  2202. were sold. Teachers liked the books, because the books encouraged 
  2203. students to read: students enjoyed reading books that contained 
  2204. their own names.
  2205.                                           Pornography (1973)
  2206.                              I programmed the computer to turn 
  2207. words into sentences. For example, Judith Ann Abrams fed the 
  2208. computer this list of words:
  2209. Names                              Other nounsTrans. verbsOther 
  2210. verbs                                                      
  2211. Adjectives                                                         
  2212. Exclamations
  2213. Barbara                            bastard bite    
  2214. disapprovebitchy                                                   
  2215. rats
  2216. Buster                             bikini  caress  fall 
  2217. downcreepy                                                         
  2218. such language
  2219. Clyde                              bitch   castratefly     dead    
  2220. ugh
  2221. Colin                              boat    climb   
  2222. hesitateenthusiastic                                             
  2223. Debbie                             cigaretteenjoy  sing    evil
  2224. Ed Shanahan                        creep   kiss    sleep   gaudy
  2225. Jeff                               dress   knife   sleepwalkgreen
  2226. Keith                              ethanol lick    sneeze  gross
  2227. Kerri                              eyelid  pick    vanish  horny
  2228. Liz                                fly     pull    walk    insane
  2229. Miles                              gamblingrip             juicy
  2230. Nicholson 700                      jerk    screw           
  2231. mixed-up
  2232. Norwalk                            knee    shake           
  2233. monstrous
  2234.                                    kneecap take off        
  2235. multicolored
  2236.                                    lip     teach           
  2237. neurotic
  2238.                                    mouth   undo            
  2239. obscene
  2240.                                    navel                   
  2241. paranoid
  2242.                                    piece                   
  2243. prickly
  2244.                                    pornography             
  2245. profane
  2246.                                    shirt                   
  2247. provocative
  2248.                                    stud                    
  2249. repulsive
  2250.                                    tongue                  
  2251. sensuous
  2252.                                    virgin                  sexual
  2253.                                    whiskey sour            sexy
  2254.                                    whore                   shiny
  2255.                                                            sick
  2256.                                                            stimul
  2257. ating
  2258.                                                            strang
  2259. e
  2260.                                                            succul
  2261. ent
  2262.                                                            tasty
  2263.                                                            unsett
  2264. led
  2265.                                                            weird
  2266. A transitive verb is an action you can do to an object. For 
  2267. example, since you can bite and caress an object, bite and caress 
  2268. are transitive verbs.
  2269.                              Here's what the computer printed:
  2270.                                     A Sexy, Profane Creep's Tongue
  2271.                                  by Judith Ann Abrams and the computer
  2272.  
  2273.                              The lip is tasty.
  2274.                              Ugh!
  2275.                              The mouth is enthusiastic. Does the 
  2276. gaudy bastard sleepwalk?
  2277.                              The tongue's whore caresses the 
  2278. bikini. When Nicholson 700 is screwing the creepy jerk over a 
  2279. piece, Jeff is teaching the gambling. Does paranoic Liz hesitate? 
  2280. Why does the tongue kiss the creep stimulatingly on horny Buster?
  2281.                              Undo! Pull the fly repulsively! A 
  2282. multicolored, enthusiastic cigarette is neurotic. Is a 
  2283. multicolored, weird dress's navel under the strange bikini? Why 
  2284. is the obscene dress's cigarette Colin's ethanol? Rats! Why is 
  2285. the dead, insane, tasty, shiny, knifing creep's bitch unsettled 
  2286. and strange?
  2287.                              Clyde is over Ed Shanahan. Kissing 
  2288. the bastard, a paranoid ethanol is unsettled. Why is the prickly 
  2289. kneecap's mouth on the creepy piece? Teaching Norwalk is sneezing 
  2290. paranoidly and grossly under provocative, shiny Debbie. As a 
  2291. sexy, walking tongue's virgin takes off the bitchy cigarette over 
  2292. the profane stud then castrates enthusiastic Mile's lip 
  2293. paranoidly, the dead, gaudy knee is stimulating and multicolored. 
  2294. Even though the dead, insane piece licks strangely and neuroticly 
  2295. in the flying, mixed-up eyelid's knee, the dead, biting, obscene 
  2296. bikini is on a repulsive mouth's gambling.
  2297.                              The pornography is gaudy. Kerri 
  2298. sleepwalks. Why is the tongue sensuous? Buster is sick. Is Miles 
  2299. monstrous? Debbie is neurotic and paranoid, when a stimulating 
  2300. fly picks the navel's jerk under Ed Shanahan. Why is the dress 
  2301. succulent? Hesitating, a kneecap sleeps and climbs the dead, 
  2302. bitchy ethanol. As insane Colin's bastard falls down weirdly in a 
  2303. sensuous dress, green, unsettled Mile's virgin is strange and 
  2304. sexual.
  2305.                              Is the creepy eyelid provocative? 
  2306. The gambling's whisky sour teaches a navel.
  2307.                              Is the gambling evil? The bitch 
  2308. walks. Is the virgin profane? Why is the navel sick? Is Liz 
  2309. enthusiastic? Debbie enjoys the creep. Fly! Shaking, green Kerri 
  2310. pulls weird Colin's fly on a sick navel, then vanishes over 
  2311. Norwalk.
  2312.                              Undo the virgin! While obscene Liz 
  2313. is juicy and sexual, profane, gaudy Jeff's knee is under a 
  2314. succulent whore's navel. Tear Keith's lip bitchily and juicily on 
  2315. sick, weird, multicolored Barbara! Why is Buster insane? The 
  2316. shirt knifes the bikini. Colin shakes the bitch. The whiskey sour 
  2317. hesitates over the green jerk. When a tasty tongue's ethanol 
  2318. walks, Kerri rips the boat and disapproves under enthusiastic 
  2319. Miles. Such language! Keith sings. Why is Buster bitchy?
  2320.                              Notice that the computer turned her 
  2321. adjectives into adverbs, by adding ly and making other changes. 
  2322. Gross became grossly, and juicy became juicily. Unfortunately, 
  2323. the computer's method wasn't perfect: the computer turned 
  2324. stimulating into stimulatingly (a non-existent word), and turned 
  2325. neurotic into neuroticly (instead of neurotically).
  2326.   It conjugated her verbs. Screw became screwing, and bite became 
  2327. biting (the computer dropped the e). Lick became licks, and teach 
  2328. became teaches (the computer added the e after the ch).
  2329.   It added 's to her nouns. Jeff became Jeff's. Miles became 
  2330. Miles's (it should have become Miles').
  2331.   For each sentence, the grammar is chosen randomly. The chance 
  2332. is 10% that the sentence will begin with an exclamation. If the 
  2333. sentence isn't merely an exclamation, the chance is 18% that the 
  2334. sentence will be a question.
  2335.   If it's a question, there's a 40% chance it will begin with the 
  2336. word why. There's a 50% chance the main part of the question will 
  2337. have the form does . . . noun phrase . . . verb phrase, and a 50% 
  2338. chance it will have this form instead: is . . . noun phrase . . . 
  2339. complement.
  2340.   To construct a noun phrase from nouns, adjectives, etc., the 
  2341. computer uses random numbers. It uses random numbers to also 
  2342. construct verb phrases and complements.
  2343.   The program uses a special variable, called W. At the beginning 
  2344. of the composition, W is near zero; but it tends to increase as 
  2345. the composition progresses. It affects the complexity. When W is 
  2346. large, the chance is large that the computer will print 
  2347. adjectives, adverbs, subordinate clauses, and correlative 
  2348. clauses.
  2349.   This sentence was produced by a small W:
  2350. The lip is tasty.
  2351. This sentence was produced by a large W:
  2352. As a sexy, walking tongue's virgin takes off the bitchy cigarette 
  2353. over the profane stud then castrates enthusiastic Mile's lip 
  2354. paranoidly, the dead, gaudy knee is stimulating and multicolored.
  2355.  
  2356.   Poetic images (1973)
  2357.   One of my students, Toby D'Oench, made the computer create 
  2358. poetic images, such as these:
  2359. TO GUINEVERE ___ LADY OF THE LAKE
  2360. Silent mists
  2361. Billow in creations
  2362. Windmills for flames evolve into ethers
  2363.   Merlin again
  2364.  
  2365.         MY MEMORY
  2366. Frozen children
  2367. Quiver with leaves
  2368. Creations with leaves hover over thoughts
  2369.   Gardens of verse
  2370.  
  2371.    A NEW ENGLAND BARN
  2372. Lazy fragrances
  2373. Waft by ethers
  2374. Seas on fragrances billow in sorrow
  2375.   Rusted pitchforks
  2376.                                  NEWPORT
  2377. Frozen sails
  2378. Slumber in fog
  2379. Hazes for sails waft by thoughts
  2380.                              Docks ___ yachts ___ luxuries of 
  2381. eras gone by
  2382.                              The program contains these lists:
  2383. Adjectives                         PrepositionsVerbs
  2384. fleeting                           of      billow in
  2385. crimson                            on      glitter with
  2386. silent                             under   flutter by
  2387. sensate                            above   drift with
  2388. pliant                             below   flow into
  2389. gloomy                             in      ponder about
  2390. pallid                             with    waft by
  2391. inky                               by      quiver with
  2392. frozen                             for     hover over
  2393. lazy                               through gleam like
  2394.                                            wander through
  2395.                                            slumber in
  2396.                                            dart by
  2397.                                            evolve into
  2398.                                            sing to
  2399.  
  2400. Title . . . noun . . . ending
  2401. TO REMBRANDT . . . windmills . . . A simple brush
  2402. WAITING FOR THE PATIENT . . . ethers . . . Waiting
  2403. THE PROPHET . . . visions . . . Then a word
  2404. LISTERINE . . . breaths . . . Plastic society
  2405. NEWPORT . . . sails . . . Docks ___ yachts ___ luxuries of eras 
  2406. gone by
  2407. EXISTENCE . . . seas . . . In the beginning?
  2408. SUMMER IN WATTS . . . flames . . . Tar-street neon ___ and the 
  2409. night
  2410. TO GUINEVERE ___ LADY OF THE LAKE . . . mists . . . Merlin again
  2411. NOON IN CALCUTTA . . . hazes . . . Emaciated dark forms strewn 
  2412. like garbage
  2413. WEST HARBOR . . . fog . . . A solitary gull slices through
  2414. A NEW ENGLAND BARN . . . fragrances . . . Rusted pitchforks
  2415. A CHILD'S MICROSCOPE . . . creations . . . The wonderful amoeba
  2416. A GROUP PORTRAIT . . . bundles . . . Christmas
  2417. THE MILKY WAY . . . cosmos . . . A gooey mess
  2418. TOMBSTONE . . . sorrow . . . Rubbings
  2419. LIFE AT THE END OF A BRANCH . . . leaves . . . Swirling to the 
  2420. ground
  2421. SEASHELLS AND THINGS . . . waves . . . Dribble-dribble-dribble 
  2422. castle
  2423. A BEAVER POND . . . reeds . . . Thwack
  2424. MY MEMORY . . . children . . . Gardens of verse
  2425. EINSTEIN . . . thoughts . . . Somehow through this ___ an 
  2426. understanding of a superior order
  2427.                              To create a poetic image, the 
  2428. computer fills in this form:
  2429.                                  TITLE
  2430. Adjective   Noun that goes with the title
  2431. Verb   Noun
  2432. Noun   Preposition   Noun   Verb   Noun
  2433.                              Ending that goes with the title
  2434.  
  2435.                                              Curses (1978)
  2436.                              Tom Dwyer & Margot Critchfield made 
  2437. the computer curse you. Here are some of the computer's curses:
  2438. May an enraged camel overwhelm your garage.
  2439. May an ancient philosopher lay an egg on your dill pickle.
  2440. May seven large chickens sing an operatic solo to your love 
  2441. letters.
  2442.                              To invent a curse, the computer 
  2443. fills in the blanks:
  2444. May   _______   ___________   your   ______.
  2445.       subject   verb phrase          object
  2446.                              The computer uses these words 
  2447. randomly:
  2448. Subjects                                 Verb phrasesObjects
  2449. an enraged camel                         send a mash note 
  2450. tomother-in-law
  2451. an ancient philosopher                   get inspiration 
  2452. frompsychoanalyst
  2453. a cocker spaniel                         redecorate  rumpus room
  2454. the Eiffel Tower                         become an obsession 
  2455. offern
  2456. a cowardly moose                         make a salt lick out 
  2457. ofgarage
  2458. the silent majority                      buy an interest inlove 
  2459. letters
  2460. the last picture show                    overwhelm   piggy bank
  2461. a furious trumpet player                 pour yogurt onhamburger
  2462. Miss America                             sing an operatic solo 
  2463. todill pickle
  2464. seven large chickens                     lay an egg onHonda
  2465.                              You can find that program on page 
  2466. 152 of their book, BASIC and the Personal Computer.
  2467.  
  2468.                  ANALYZE WRITING
  2469.   The computer can analyze what humans write.
  2470.  
  2471.                  English poetry
  2472.   Can the computer analyze English poetry? From 1957 to 1959 at 
  2473. Cornell University, Stephen Parrish made the computer alphabetize 
  2474. the words in Matthew Arnold's poetry. Here's an excerpt:
  2475.                                                       Page                 
  2476. Line
  2477.                                                        in                   
  2478. in
  2479.                                                       book  
  2480. Poem's title   poem
  2481. CONSCIOUS
  2482.   back with the conscious thrill of shame              181  
  2483. Isolation Marg  19
  2484.   conscious or not of the past                         287  Rugby 
  2485. Chapel    45
  2486. CONSCIOUSNESS
  2487.   the last spark of man's consciousness with words     429  
  2488. Empedocles II   30
  2489.   and keep us prisoners of our consciousness           439  
  2490. Empedocles II  352
  2491. CONSECRATE
  2492.   Peter his friend with light did consecrate           445  
  2493. Westmin Abbey   50
  2494. CONSECRATES
  2495.   which consecrates the ties of blood for these indeed 196  Frag 
  2496. Antigone   31
  2497. CONSECRATION
  2498.   won consecration from time                           281  
  2499. Haworth Church  46
  2500.   foreshown thee in thy consecration-hour              446  
  2501. Westmin Abbey   75
  2502. To find out what Matthew Arnold said about love, just look up 
  2503. LOVE. Such an index is called a concordance.
  2504.   That concordance was the first produced by a computer. 
  2505. Previously, all concordances of poetry were created by hand, 
  2506. using filing cards. For example, in 1870 a group of researchers 
  2507. began creating a concordance to Chaucer, by hand. They started at 
  2508. the letter A. 45 years later, they were only up to the letter H!
  2509.   Did the poet Shelley steal ideas from others? Joseph Raben, at 
  2510. Queens College, believed Shelley borrowed imagery from Milton. To 
  2511. prove it, in 1964 he made the computer produce concordances to 
  2512. Shelley's Prometheus Unbound and Milton's Paradise Lost and 
  2513. compare them. The computer found many similarities between 
  2514. Shelley and Milton.
  2515.   What were Shakespeare's favorite words? In 1971 at Münster 
  2516. University in Germany, Marvin Spevack fed the computer all the 
  2517. works of Shakespeare, and made it count how often each word 
  2518. occurs. Disregarding trivial words such as a and the, the 
  2519. computer discovered Shakespeare's favorite word was love: he used 
  2520. it 2,271 times. Next come heart, death, man, life, and hand. He 
  2521. never used the word hero. In Macbeth, the word good occurs more 
  2522. often than any other adjective, noun, or adverb, and more often 
  2523. than most verbs.
  2524.   By counting words, other researchers made the computer graph 
  2525. the rise and fall of themes in a novel.
  2526.  
  2527.                 American history
  2528.   Who wrote the Federalist Papers? Historians knew some of the 
  2529. papers were by Alexander Hamilton and others by James Madison, 
  2530. but the authorship of the remaining papers was in dispute.
  2531.   In 1964, Mosteller and Wallace made the computer compare the 
  2532. literary styles of the papers, by counting the frequency of words 
  2533. such as by, enough, from, to, upon, while, and whilst. It 
  2534. concluded that all the disputed papers were written by Madison, 
  2535. not Hamilton.
  2536.   The statistical evidence was so high that historians accept the 
  2537. computer's finding as fact.
  2538.  
  2539.                     The Bible
  2540.   Can the computer analyze the Bible? In 1951, Texas clergyman 
  2541. John Ellison made the computer compare 309 Greek manuscripts of 
  2542. the New Testament. Underneath each word of a standard text, the 
  2543. computer printed the variants found in other manuscripts. It 
  2544. classified the manuscripts according to their similarities.
  2545.   In 1957, he published a concordance to the Revised Standard 
  2546. Bible, and a pair of other researchers (Tasman & Busa) indexed 
  2547. the Dead Sea Scrolls.
  2548.                                                      Did the 
  2549. apostle Paul really write all those marvelous letters attributed 
  2550. to him in the New Testament? Or were they actually written by 
  2551. somebody else?
  2552.                                                      In 1964, 
  2553. Scottish clergyman Andrew Morton used the computer to deduce that 
  2554. Paul didn't write some of those letters.
  2555.                                                      All Morton 
  2556. did was count how often Paul used the Greek word kai in each 
  2557. sentence. Kai means and. Coming to a conclusion about Biblical 
  2558. authorship by counting just the word and might seem silly, but 
  2559. Morton said he analyzed 20 writers of ancient Greek and found 
  2560. each used kai with a constant frequency. In the ``Pauline'' 
  2561. letters, the frequency of kai varied a lot, implying some of them 
  2562. were not by Paul.
  2563.                                                      Ellison 
  2564. distrusted Morton's assumption that a man's literary style must 
  2565. remain constant. He warned: if Morton's method were applied to 
  2566. the Declaration of Independence and Thomas Jefferson's letters to 
  2567. his wife, the computer might conclude that either Jefferson 
  2568. didn't write the Declaration of Independence or another man was 
  2569. writing love letters to Mrs. Jefferson. In 1965, to prove his 
  2570. point, he applied Morton's method to two of Morton's own articles 
  2571. on the subject: the computer concluded that Morton could not be 
  2572. the author of both!
  2573.  
  2574.                                                             Forgery
  2575.                                                      IBM 
  2576. programmed the computer to detect a forged signature ___ even if 
  2577. the signature looks correct to the naked eye.
  2578.                                                      To use the 
  2579. IBM forgery-detection system, write your signature by using IBM's 
  2580. special pen, attached to the computer. As you write, the computer 
  2581. notices how hard you press the pen against the paper and how fast 
  2582. you move the pen.
  2583.                                                      If somebody 
  2584. else tries to pretend he's you, he must sit down at the machine 
  2585. and try to duplicate your signature. If he presses the pen 
  2586. hardest at different points of the signature, or if he 
  2587. accelerates the pen's motion at different points, the computer 
  2588. says he's a fake.
  2589.                                                      The system 
  2590. works well, because the average crook trying to forge your 
  2591. signature will hesitate at the hard parts. His hesitation affects 
  2592. the pen's pressure and acceleration, which tell the computer he's 
  2593. faking.
  2594.                                                      IBM 
  2595. developed the system in 1979 but isn't selling it yet. When it 
  2596. does, remember: the system works just on signatures written with 
  2597. IBM's pen.
  2598.  
  2599.           TRANSLATE RUSSIAN
  2600.   Soon after computers were invented, programmers tried to make 
  2601. them translate Russian into English. They chose Russian instead 
  2602. of Spanish, for three reasons:
  2603.   1. Few humans could translate Russian. Spanish translators were 
  2604. a-dime-a-dozen.
  2605.   2. Computer experts love hard problems. Russian is harder than 
  2606. Spanish.
  2607.   3. Most computers were owned by the Department of Defense, 
  2608. which is very interested in Russia.
  2609.  
  2610.            Early attempts
  2611.   In 1954, IBM wrote a program that translated Russian sentences 
  2612. such as:
  2613. Gasoline is prepared by chemical methods from crude oil. The 
  2614. price of crude oil is determined by the market. The quality of 
  2615. the crude oil is determined by the calorie content.
  2616.   Unfortunately, most Russian sentences are not so simple. During 
  2617. the 1960's, the end of a Russian paper on space biology was fed 
  2618. into an advanced program written by Computer Concepts, Inc. 
  2619. Here's the translation that came out:
  2620. Thus, the examination of some from fundamental RADIOBIOLOGICESKIX 
  2621. problems shows, that in this a field still very much NEREWENNYZ 
  2622. questions. This is clear, since cosmic RADIOBIOLOGI4 is very 
  2623. young RAZDELOM young science efforts of scientific different 
  2624. specialties of the different countries of the world successful 
  2625. PRODOLJENY will be expanded there are.
  2626. The computer couldn't translate the words in capital letters and 
  2627. was stumped by Russian grammar.
  2628.   The competing program, written by the Air Force, translated the 
  2629. same passage a little better:
  2630. Thus, consideration of from basic radio-biological problems shows 
  2631. that in a given region still very many unsolved questions. This 
  2632. and intelligibly, since space radiobiology is very young division 
  2633. of young science ___ space biology. However, is base to trust 
  2634. that jointly scientists of different specialties of various 
  2635. countries of world/peace radiobiological investigations in outer 
  2636. space will be successfully continued and expanded.
  2637.   In 1966, a special committee of the National Academy of 
  2638. Sciences concluded that the experience of computer translation 
  2639. was ``uniformly discouraging'' and that hiring a human translator 
  2640. was cheaper than doing the two-step process of computer 
  2641. translation followed by human editing.
  2642.   During the last 20 years, computer prices have fallen, but so 
  2643. has the availability of Americans who know Russian, so the 
  2644. computer's usefulness is still in doubt. Today, most translations 
  2645. are still done by humans, who use computers to help do the word 
  2646. processing and to search through a dictionary and thesaurus.
  2647.                                                    Famous errors
  2648.                                          If you program the 
  2649. computer to translate an English sentence into Russian, and then 
  2650. the Russian back to English, will you get back the same English 
  2651. sentence you started with?
  2652.                                          One programmer tried, 
  2653. ``The spirit is willing, but the flesh is weak.'' The computer 
  2654. translated it into Russian, then back into English, and printed, 
  2655. ``The booze is strong, but the meat is rotten.''
  2656.                                          Another programmer 
  2657. tried, ``Out of sight, out of mind.'' The computer printed, 
  2658. ``Blind idiot.''
  2659.                                          At an engineering 
  2660. conference, a computer was translating scientific papers into 
  2661. English, when it suddenly started talking about ``water sheep''. 
  2662. Everyone was confused. Finally they figured it out: the computer 
  2663. meant hydraulic rams.
  2664.  
  2665.                                         Xerox's amazing translation machine
  2666.                                          In Moscow during the 
  2667. 1960's, American companies were showing off their products, but 
  2668. none of the Russians were interested in Xerox's photocopiers ___ 
  2669. until some Xerox employees put on an amazing demonstration. They 
  2670. ``photocopied'' some English writing, and ___ presto! ___ a 
  2671. beautiful Russian translation of it came out of the machine! The 
  2672. machine was acting as a translator! And the translation was 
  2673. flawless, even though the English text was complex!
  2674.                                          The Russians were very 
  2675. excited, and ordered hundreds of the amazing translation machine.
  2676.                                          But before shipping the 
  2677. machines, the Xerox guys confessed it was just a gag. The 
  2678. employees had sneaked the Russian version into the machine, 
  2679. before beginning the demonstration.
  2680.                                          What if Americans had 
  2681. the same sense of humor about nuclear war? ``Hello, Gorbachev? 
  2682. This is George Bush, on the hot line. We just fired some nuclear 
  2683. missiles. They're heading straight for Moscow. Ha, ha! Just 
  2684. kidding.''
  2685. Verbs
  2686. moved
  2687. drifted
  2688. shone
  2689. floated
  2690. touched
  2691. melted
  2692. looked down on
  2693. warmed
  2694.  
  2695. Verbs
  2696. scampered
  2697. flew
  2698. ran
  2699. hurried
  2700. sniffed at
  2701. peered at
  2702. ate
  2703. munched and crunched
  2704.