1
00:00:01,920 --> 00:00:06,840
Speaker 1: Welcome to Brainstuff production of I Heart Radio, Hey brain

2
00:00:06,920 --> 00:00:10,920
Speaker 1: Stuff Lauren Vogelbaum here. The Turning test is legendary in

3
00:00:10,920 --> 00:00:14,360
Speaker 1: the field of artificial intelligence. First proposed by the visionary

4
00:00:14,360 --> 00:00:17,760
Speaker 1: British mathematician Alan Turning in a landmark nineteen fifty paper.

5
00:00:18,120 --> 00:00:20,880
Speaker 1: The test provides a practical and pretty fun way to

6
00:00:20,920 --> 00:00:24,160
Speaker 1: determine if a computer has achieved human levels of intelligence.

7
00:00:24,640 --> 00:00:28,240
Speaker 1: Turing called it the imitation game. If a computer, through

8
00:00:28,280 --> 00:00:30,960
Speaker 1: a text only chat, can convince a human that it's

9
00:00:30,960 --> 00:00:34,320
Speaker 1: a real person, then it passes the test. Simple in

10
00:00:34,400 --> 00:00:38,280
Speaker 1: theory but nearly impossible in practice. Turning came up with

11
00:00:38,360 --> 00:00:40,760
Speaker 1: the imitation game in response to colleagues and critics in

12
00:00:40,800 --> 00:00:43,640
Speaker 1: the late nineteen forties who insisted that a machine could

13
00:00:43,680 --> 00:00:47,360
Speaker 1: never be truly intelligent. But Turning had more faith in

14
00:00:47,440 --> 00:00:51,520
Speaker 1: these primitive new machines he called digital computers. That's because

15
00:00:51,520 --> 00:00:53,680
Speaker 1: Turing was the very first to envision something that we

16
00:00:53,760 --> 00:00:56,480
Speaker 1: take for granted today, a single machine that can be

17
00:00:56,520 --> 00:01:00,320
Speaker 1: programmed to do almost anything. Odds are yours sing to

18
00:01:00,320 --> 00:01:04,440
Speaker 1: this podcast on just such a machine? In brief, Alan

19
00:01:04,520 --> 00:01:06,800
Speaker 1: Turing was a British mathematician who came up with the

20
00:01:06,840 --> 00:01:09,640
Speaker 1: idea of modern computing, and whose code breaking played a

21
00:01:09,680 --> 00:01:12,000
Speaker 1: major role in the Allied victory over the Nazis in

22
00:01:12,040 --> 00:01:15,280
Speaker 1: World War Two. Also, he was a world class cross

23
00:01:15,280 --> 00:01:17,880
Speaker 1: country runner who may have qualified for the nineteen forty

24
00:01:18,000 --> 00:01:21,000
Speaker 1: Olympics if not for an injury. But his life was

25
00:01:21,040 --> 00:01:24,240
Speaker 1: also tragic due to prejudices of the time. He was

26
00:01:24,280 --> 00:01:26,600
Speaker 1: prosecuted in nineteen fifty two for having an affair with

27
00:01:26,640 --> 00:01:29,720
Speaker 1: another man home sexual acts being illegal in Britain until

28
00:01:29,840 --> 00:01:32,880
Speaker 1: nineteen sixty seven, and he accepted a form of chemical

29
00:01:32,920 --> 00:01:35,720
Speaker 1: castration as a condition of probation in order to avoid

30
00:01:35,840 --> 00:01:39,520
Speaker 1: jail time. His security clearance was revoked, ending his work

31
00:01:39,520 --> 00:01:42,080
Speaker 1: for the British government. He was found dead of cyanid

32
00:01:42,120 --> 00:01:45,200
Speaker 1: poisoning in nineteen fifty four, though it's still unclear whether

33
00:01:45,200 --> 00:01:48,760
Speaker 1: his death was a suicide or an accident. He was

34
00:01:48,800 --> 00:01:53,080
Speaker 1: pardoned of his conviction by Queen Elizabeth. The second Turing

35
00:01:53,160 --> 00:01:56,160
Speaker 1: was writing about computers well before any such thing existed.

36
00:01:56,640 --> 00:01:59,520
Speaker 1: Back in nineteen thirty six, he introduced the concept of

37
00:01:59,560 --> 00:02:02,760
Speaker 1: the univer soul computing machine and a dense mathematical paper

38
00:02:03,040 --> 00:02:07,840
Speaker 1: called on Computable Numbers with an application to the chitons problem.

39
00:02:07,880 --> 00:02:10,519
Speaker 1: This was a decade before the first electronic computer would

40
00:02:10,560 --> 00:02:14,480
Speaker 1: be built. Turning wrote, according to my definition, a number

41
00:02:14,560 --> 00:02:17,120
Speaker 1: is computable if its decimal can be written down by

42
00:02:17,160 --> 00:02:20,079
Speaker 1: a machine. It is possible to invent a single machine

43
00:02:20,120 --> 00:02:24,679
Speaker 1: which can be used to compute any computable sequence. Turing's

44
00:02:24,680 --> 00:02:27,720
Speaker 1: definition of computability of something that a computer can do

45
00:02:28,160 --> 00:02:31,519
Speaker 1: is what's known today as an algorithm. Turning was the

46
00:02:31,560 --> 00:02:33,920
Speaker 1: first to lay out the design framework of a machine

47
00:02:34,080 --> 00:02:36,600
Speaker 1: that could be programmed to run a series of discrete

48
00:02:36,639 --> 00:02:40,959
Speaker 1: algorithms in order to achieve a desired task. Other mathematicians

49
00:02:40,960 --> 00:02:44,680
Speaker 1: and engineers had toyed with calculating machines, most famously Charles

50
00:02:44,720 --> 00:02:48,880
Speaker 1: Babbage's nineteenth century analytical engine, but Turing envisioned a device

51
00:02:48,919 --> 00:02:52,960
Speaker 1: that wasn't limited to solving one kind of problem. We

52
00:02:53,000 --> 00:02:56,480
Speaker 1: spoke with Andrew Hodges, a mathematics professor at Oxford University

53
00:02:56,639 --> 00:03:00,360
Speaker 1: an author of Alan Turing The Enigma, the inspiration for

54
00:03:00,400 --> 00:03:05,680
Speaker 1: the Oscar winning film The Imitation Game. Hodges explained, anything

55
00:03:05,720 --> 00:03:08,079
Speaker 1: you can describe as an algorithm can be done by

56
00:03:08,120 --> 00:03:11,480
Speaker 1: one machine. The universal machine is essentially what we mean

57
00:03:11,480 --> 00:03:14,080
Speaker 1: by a computer, now, something on which you can store

58
00:03:14,120 --> 00:03:17,000
Speaker 1: the instructions and it carries them out, and no one

59
00:03:17,040 --> 00:03:21,840
Speaker 1: else had formalized that idea. From the start, Turns universal

60
00:03:21,880 --> 00:03:25,519
Speaker 1: machine was conceived as a very simplified form of artificial intelligence,

61
00:03:25,840 --> 00:03:28,359
Speaker 1: even though that term wouldn't be coined until ninety six.

62
00:03:29,240 --> 00:03:31,840
Speaker 1: Hodges says that the design of the universal machine was

63
00:03:31,880 --> 00:03:34,240
Speaker 1: meant to imitate the inner workings of the human mind,

64
00:03:34,520 --> 00:03:37,520
Speaker 1: a subject that fascinated Turing almost as much as mathematics.

65
00:03:38,280 --> 00:03:41,040
Speaker 1: In fact, when describing how his universal machine would work,

66
00:03:41,280 --> 00:03:43,880
Speaker 1: Turn used the term state of mind to label the

67
00:03:43,920 --> 00:03:47,720
Speaker 1: different read and write functions of the machine. In Turn's

68
00:03:47,760 --> 00:03:50,440
Speaker 1: conceptual machine, a length of tape is run through a

69
00:03:50,480 --> 00:03:53,440
Speaker 1: read write scanner. The tape is inscribed with bits of

70
00:03:53,480 --> 00:03:57,560
Speaker 1: information represented by symbols. The scanner head can either read

71
00:03:57,600 --> 00:04:00,400
Speaker 1: the symbols or write new ones according to its state

72
00:04:00,440 --> 00:04:04,040
Speaker 1: of mind. Turning wrote in his nineteen thirty six paper,

73
00:04:04,480 --> 00:04:07,600
Speaker 1: the operation actually performed is determined by the state of

74
00:04:07,640 --> 00:04:11,080
Speaker 1: mind of the computer and the observed symbols. In particular,

75
00:04:11,240 --> 00:04:13,560
Speaker 1: they determine the state of mind of the computer after

76
00:04:13,720 --> 00:04:17,960
Speaker 1: the operation is carried out. A decade later, when Turing

77
00:04:18,000 --> 00:04:20,000
Speaker 1: was leading the stalled British effort to build one of

78
00:04:20,000 --> 00:04:23,160
Speaker 1: the first electronic computers in nineteen forty six. He also

79
00:04:23,160 --> 00:04:26,839
Speaker 1: studied neurology and human physiology on the side. The result

80
00:04:26,880 --> 00:04:29,920
Speaker 1: was an internal paper published for the National Physical Laboratory

81
00:04:30,000 --> 00:04:32,679
Speaker 1: that modeled how a computer could be programmed to learn

82
00:04:32,880 --> 00:04:35,560
Speaker 1: on its own. Hod Just sees it as one of

83
00:04:35,560 --> 00:04:38,479
Speaker 1: the earliest proposals of what are now called neural networks,

84
00:04:38,720 --> 00:04:41,080
Speaker 1: a type of deep machine learning that's at the bleeding

85
00:04:41,160 --> 00:04:45,719
Speaker 1: edge of artificial intelligence. Turing wasn't the only person intrigued

86
00:04:45,760 --> 00:04:49,120
Speaker 1: by the similarities between human and machine intelligence. A surge

87
00:04:49,120 --> 00:04:52,839
Speaker 1: of new technologies developed during World War Two, including early computers,

88
00:04:52,880 --> 00:04:56,240
Speaker 1: space satellites, and nuclear power, had captured the intellectual and

89
00:04:56,400 --> 00:05:00,240
Speaker 1: public imagination. Hodges said, as soon as computer as are

90
00:05:00,240 --> 00:05:03,400
Speaker 1: mentioned at all, people are talking about electronic brains and

91
00:05:03,400 --> 00:05:07,800
Speaker 1: the possibility of the computer rivaling the brain. The nineteen

92
00:05:07,839 --> 00:05:10,960
Speaker 1: forty eight books Cybernetics by Norbert Wiener coined the prefix

93
00:05:11,000 --> 00:05:13,520
Speaker 1: cyber and wondered whether it would be possible to quote

94
00:05:13,560 --> 00:05:16,279
Speaker 1: construct a chess playing machine and whether this sort of

95
00:05:16,279 --> 00:05:19,279
Speaker 1: ability represents an essential difference between the potentialities of the

96
00:05:19,279 --> 00:05:23,520
Speaker 1: machine and the mind. Wiener concluded that such a machine

97
00:05:23,880 --> 00:05:25,680
Speaker 1: might very well be as good a player as the

98
00:05:25,760 --> 00:05:29,560
Speaker 1: vast majority of the human race. It was during this

99
00:05:29,680 --> 00:05:33,279
Speaker 1: era of excitement and nervous speculation about superintelligent machines the

100
00:05:33,320 --> 00:05:37,039
Speaker 1: Turning wrote Computing, Machinery and Intelligence, what Hodges calls one

101
00:05:37,080 --> 00:05:41,440
Speaker 1: of the most cited papers in philosophical literature. Turning begins,

102
00:05:41,760 --> 00:05:46,360
Speaker 1: I propose to consider the question can machines think? Then,

103
00:05:46,480 --> 00:05:49,400
Speaker 1: since the definitions of machine and think are ambiguous, Turning

104
00:05:49,520 --> 00:05:52,440
Speaker 1: narrows the scope of the question for his purposes. The

105
00:05:52,440 --> 00:05:55,159
Speaker 1: machine must be a digital computer, and the test of

106
00:05:55,160 --> 00:05:57,560
Speaker 1: whether or not it can think would be answered by

107
00:05:57,640 --> 00:06:00,839
Speaker 1: the imitation game. The game game, now known as the

108
00:06:00,839 --> 00:06:03,800
Speaker 1: Turning Test, is only mentioned briefly in the paper, and

109
00:06:03,880 --> 00:06:05,919
Speaker 1: Hodges says that Turing didn't take the details of the

110
00:06:05,920 --> 00:06:09,599
Speaker 1: test too seriously, publishing different versions in other papers. But

111
00:06:09,680 --> 00:06:13,360
Speaker 1: Turning did like the playful simplicity of it. Hodges said,

112
00:06:13,720 --> 00:06:15,919
Speaker 1: in a way he was making a drama out of it.

113
00:06:15,920 --> 00:06:18,720
Speaker 1: It presented this idea of the possibility of advanced artificial

114
00:06:18,800 --> 00:06:21,560
Speaker 1: intelligence in a way that engages people, and that ordinary

115
00:06:21,560 --> 00:06:24,120
Speaker 1: people would make the decision like a jury in a trial.

116
00:06:25,440 --> 00:06:27,840
Speaker 1: When the Turing Test was first published in nineteen fifty

117
00:06:27,960 --> 00:06:31,160
Speaker 1: Turning himself was confident that intelligent machinery, as he called it,

118
00:06:31,320 --> 00:06:33,760
Speaker 1: would be able to win the Invitation Game within fifty

119
00:06:33,800 --> 00:06:37,400
Speaker 1: to a hundred years. So will his predictions come true?

120
00:06:38,320 --> 00:06:41,440
Speaker 1: We already have super intelligent computers capable of outwitting the

121
00:06:41,440 --> 00:06:46,040
Speaker 1: smartest players in other types of games. IBM's Deep Blue

122
00:06:46,080 --> 00:06:49,640
Speaker 1: defeated the reigning chess champion Gary Kasparov, and Watson beat

123
00:06:49,680 --> 00:06:52,920
Speaker 1: the record breaking Jeopardy champion Ken Jennings in two thousand eleven.

124
00:06:54,040 --> 00:06:57,680
Speaker 1: But the Imitation Game raises the bar high on artificial intelligence,

125
00:06:57,880 --> 00:07:00,479
Speaker 1: and no computer has come close to convincing an ordinary

126
00:07:00,560 --> 00:07:03,520
Speaker 1: human that is one of us, at least not yet.

127
00:07:04,560 --> 00:07:07,440
Speaker 1: An annual contest called the Loebner Prize conducts its own

128
00:07:07,480 --> 00:07:09,720
Speaker 1: turing tests on the top chat bots to see if

129
00:07:09,760 --> 00:07:12,280
Speaker 1: the latest AI software could convince a panel of judges

130
00:07:12,440 --> 00:07:15,880
Speaker 1: that it's just as human as its human creators. None

131
00:07:15,880 --> 00:07:18,880
Speaker 1: of the chat bots have succeeded. The best performer, a

132
00:07:18,920 --> 00:07:22,200
Speaker 1: conversational chatbot called Mistaku, has only achieved a rating of

133
00:07:22,320 --> 00:07:26,200
Speaker 1: thirty three human, but when our writer Dave went online

134
00:07:26,200 --> 00:07:28,280
Speaker 1: to chat with her, he was impressed by her natural

135
00:07:28,360 --> 00:07:31,800
Speaker 1: language responses and deep knowledge, albeit probably too deep for

136
00:07:31,800 --> 00:07:35,360
Speaker 1: a typical dope human, he said. And when he asked

137
00:07:35,360 --> 00:07:37,480
Speaker 1: her if a chat bot will ever pass the Turing test,

138
00:07:37,600 --> 00:07:40,360
Speaker 1: she had the perfect answer, you be the judge of that.

139
00:07:45,640 --> 00:07:48,520
Speaker 1: Today's episode written by Dave Ruse and produced by Tyler Clang.

140
00:07:48,920 --> 00:07:50,760
Speaker 1: Brain Stuff is a production of I Heart Radios How

141
00:07:50,760 --> 00:07:52,680
Speaker 1: Stuff Works. For more in this and lots of other

142
00:07:52,720 --> 00:07:55,960
Speaker 1: intelligent topics, visit our home planet, how stuff Works dot com.

143
00:07:56,000 --> 00:07:58,240
Speaker 1: And for more podcasts for My heart Radio, including one

144
00:07:58,240 --> 00:08:01,080
Speaker 1: from Ken Jennings called Omnibus, visit that I Heart Radio app,

145
00:08:01,200 --> 00:08:03,760
Speaker 1: app podcasts, or wherever you listen to your favorite shows.

146
00:08:14,760 --> 00:08:14,800
Speaker 1: H