1
00:00:00,760 --> 00:00:03,840
Speaker 1: Hi everyone, it's Josh and for this week's select, I've

2
00:00:03,920 --> 00:00:08,479
Speaker 1: chosen our twenty eighteen episode Some Games You would Surely

3
00:00:08,560 --> 00:00:13,160
Speaker 1: Lose to a Computer. It's a philosophical discussion about AI

4
00:00:13,280 --> 00:00:16,880
Speaker 1: that's disguised as an episode on computer games. Honestly, we

5
00:00:16,920 --> 00:00:19,040
Speaker 1: didn't plan it to be like that. It just turned

6
00:00:19,040 --> 00:00:21,960
Speaker 1: out that way. We're pretty happy that it did. And

7
00:00:22,000 --> 00:00:24,480
Speaker 1: in light of the recent advances with machine learning like

8
00:00:24,800 --> 00:00:27,520
Speaker 1: chat GPT, a few of the things we say seemed

9
00:00:27,680 --> 00:00:30,479
Speaker 1: naively quaint now. Plus it has a doll up of

10
00:00:30,480 --> 00:00:34,080
Speaker 1: our tech stuff colleague Jonathan stricklanet and so that's a bonus.

11
00:00:34,320 --> 00:00:36,720
Speaker 1: I hope you enjoy.

12
00:00:39,479 --> 00:00:45,240
Speaker 2: Welcome to Stuff you should know, a production of iHeartRadio.

13
00:00:49,240 --> 00:00:52,440
Speaker 1: Hey, and welcome to the podcast. I'm Josh Clark. There's

14
00:00:52,520 --> 00:00:56,840
Speaker 1: Charles w Chuck Bryant, there's Jerry over there. I'm just

15
00:00:56,880 --> 00:00:58,960
Speaker 1: gonna come out and tell everybody making fun of me

16
00:00:59,080 --> 00:01:04,559
Speaker 1: for some weird reasons, vaguely weird ways. But I'm all right,

17
00:01:05,040 --> 00:01:07,840
Speaker 1: So check out his story for you. Okay, I'm going

18
00:01:07,920 --> 00:01:11,880
Speaker 1: to take us back to the seventeen seventies and they'll

19
00:01:12,080 --> 00:01:17,520
Speaker 1: swing in town of Vienna, not Virginia, not Vanna Georgia,

20
00:01:18,120 --> 00:01:20,800
Speaker 1: which you know, that's how they pronounce it, right, Yanna,

21
00:01:21,280 --> 00:01:24,440
Speaker 1: Vienna sausages, right, Vienna Austria.

22
00:01:24,480 --> 00:01:25,240
Speaker 3: Have you ever been there?

23
00:01:26,080 --> 00:01:30,000
Speaker 1: Vienna, Austria. No, been to Brussels. That was pretty close.

24
00:01:30,720 --> 00:01:32,520
Speaker 3: Vienna's lovely, I'm sure.

25
00:01:32,880 --> 00:01:34,560
Speaker 1: I think it's a lot like Brussels.

26
00:01:35,319 --> 00:01:38,800
Speaker 3: Very clean, lovely town. I just remember it being very clean.

27
00:01:39,240 --> 00:01:44,040
Speaker 1: Yeah, very clean, gorgeous architecture, weird little angled side streets.

28
00:01:44,040 --> 00:01:48,600
Speaker 1: They're very narrow, very pretty town. So we're in Vienna

29
00:01:48,680 --> 00:01:52,840
Speaker 1: and there is a dude skulking about going to the

30
00:01:52,960 --> 00:01:58,680
Speaker 1: royal Palace in Vienna. His name is Wolfgang von Kempelin,

31
00:01:59,600 --> 00:02:02,560
Speaker 1: and he's he's an inventor, he's an engineer. He's a

32
00:02:02,600 --> 00:02:05,320
Speaker 1: pretty sharp dude. And he's got with him what would

33
00:02:05,320 --> 00:02:08,480
Speaker 1: come to be known as the Turk, but he called

34
00:02:08,520 --> 00:02:13,480
Speaker 1: it the mechanical Turk or the automaton chess player, and

35
00:02:13,480 --> 00:02:17,560
Speaker 1: that's what it was. It was a wooden figure that

36
00:02:17,760 --> 00:02:21,959
Speaker 1: moved mechanically, seated at a cabinet, and on top of

37
00:02:22,000 --> 00:02:25,360
Speaker 1: the cabinet was a chessboard. And when he brought it

38
00:02:25,400 --> 00:02:30,000
Speaker 1: out to show to the royal court, he would it

39
00:02:30,040 --> 00:02:33,919
Speaker 1: was cool kind of but nothing they hadn't seen before,

40
00:02:34,000 --> 00:02:38,000
Speaker 1: because automata was kind of a hip thing by then.

41
00:02:38,760 --> 00:02:45,400
Speaker 3: Yeah, people loved building these engineering these automata machines to

42
00:02:45,440 --> 00:02:48,359
Speaker 3: do various things. And people are just knocked out by

43
00:02:48,480 --> 00:02:51,160
Speaker 3: the fact that, you know, you hide these gears and

44
00:02:51,240 --> 00:02:54,960
Speaker 3: levers behind wood or a cloth, and it looks as

45
00:02:54,960 --> 00:02:57,640
Speaker 3: though there's a real well not real, but you know.

46
00:02:57,639 --> 00:02:59,919
Speaker 1: What I mean that it's like a real machine.

47
00:03:00,600 --> 00:03:02,880
Speaker 3: Yeah, but not. They weren't fooled anything like is that

48
00:03:02,960 --> 00:03:06,760
Speaker 3: a real man? It was, but it was for their time.

49
00:03:06,880 --> 00:03:09,920
Speaker 3: It was so advanced looking that it's like us seeing

50
00:03:10,639 --> 00:03:12,720
Speaker 3: x Makina in the movie theater.

51
00:03:12,960 --> 00:03:15,760
Speaker 1: Sure does that make sense? Yeah, no, it does make sense.

52
00:03:15,800 --> 00:03:19,400
Speaker 1: But imagine seeing like X makinan being like I've seen

53
00:03:19,440 --> 00:03:22,160
Speaker 1: this before. This isn't anything special, okay.

54
00:03:22,440 --> 00:03:24,520
Speaker 3: Yeah, And this thing, to be clear, looked like a

55
00:03:25,520 --> 00:03:30,799
Speaker 3: is it Zoltar or Zultan from Big Zoltar Zultan?

56
00:03:31,360 --> 00:03:32,680
Speaker 1: I don't know. It's one of those two.

57
00:03:32,720 --> 00:03:34,560
Speaker 3: One of those two. Like this, this guy's wearing a

58
00:03:34,600 --> 00:03:39,440
Speaker 3: turban and it's in a glass case like a bust like,

59
00:03:39,480 --> 00:03:41,040
Speaker 3: you know, like a chest up thing.

60
00:03:41,160 --> 00:03:44,080
Speaker 1: Yeah, he's seated at this cabinet, so there's no need

61
00:03:44,120 --> 00:03:45,520
Speaker 1: for legs or anything like that.

62
00:03:45,680 --> 00:03:45,960
Speaker 3: Yeah.

63
00:03:46,040 --> 00:03:49,320
Speaker 1: But the thing, this is what was amazing about the Turk.

64
00:03:50,240 --> 00:03:53,160
Speaker 1: He could play chess, and he could play chess really well.

65
00:03:54,000 --> 00:03:57,240
Speaker 1: So yeah, he was like an automaton and he moved

66
00:03:57,280 --> 00:04:00,880
Speaker 1: all herky jerky or whatever. But he could play you

67
00:04:00,960 --> 00:04:07,120
Speaker 1: in chess, which was a huge, huge advance at the time. Like,

68
00:04:07,160 --> 00:04:10,240
Speaker 1: this is something that wouldn't come up again until the

69
00:04:10,320 --> 00:04:13,720
Speaker 1: nineteen nineties, more than two hundred years later. This thing,

70
00:04:13,960 --> 00:04:17,080
Speaker 1: this automaton could play a human being in chess and

71
00:04:17,200 --> 00:04:17,839
Speaker 1: beat them.

72
00:04:18,120 --> 00:04:20,520
Speaker 3: Well. Yeah, and it looked like when the game started,

73
00:04:20,560 --> 00:04:22,960
Speaker 3: it would look down at the chessboard and like cock

74
00:04:23,080 --> 00:04:25,240
Speaker 3: his head, like what should my first move be?

75
00:04:25,520 --> 00:04:25,760
Speaker 1: Right?

76
00:04:26,040 --> 00:04:28,520
Speaker 3: And if people I love this part. If people tried

77
00:04:28,560 --> 00:04:32,080
Speaker 3: to cheat. Apparently Napoleon tried to cheat this thing because

78
00:04:32,120 --> 00:04:35,479
Speaker 3: this guy he debuted at the VI and East Court.

79
00:04:35,960 --> 00:04:37,880
Speaker 3: But then it, you know, it went on a world tour.

80
00:04:38,200 --> 00:04:41,000
Speaker 1: Yeah, and he was even it was taken over by

81
00:04:41,040 --> 00:04:43,080
Speaker 1: a successor to the guy who toured with it.

82
00:04:43,120 --> 00:04:44,920
Speaker 3: Even further, people went nuts for this stuff.

83
00:04:44,920 --> 00:04:46,839
Speaker 1: They did. They loved it because they were like, this

84
00:04:46,920 --> 00:04:49,840
Speaker 1: is crazy, I can't believe what I'm seeing. Most people, though,

85
00:04:49,880 --> 00:04:51,640
Speaker 1: were not taken in by it, they're like, there's some

86
00:04:51,720 --> 00:04:54,880
Speaker 1: trick here. Sure, but von Kempelin and the guy who

87
00:04:54,920 --> 00:04:58,200
Speaker 1: came after him, I don't remember his name, they would

88
00:04:58,240 --> 00:05:00,960
Speaker 1: demonstrate you could open this happened and you could see

89
00:05:01,000 --> 00:05:04,960
Speaker 1: all the workings of the mechanical turk inside.

90
00:05:05,600 --> 00:05:08,280
Speaker 3: Right, So what I was saying is if this thing's

91
00:05:08,560 --> 00:05:13,400
Speaker 3: since a cheater like Napoleon supposedly did it, would you

92
00:05:13,440 --> 00:05:15,880
Speaker 3: know Napoleon would move a piece out of turner illegally

93
00:05:15,960 --> 00:05:19,040
Speaker 3: or something. This dude, the turk Turk one eighty two

94
00:05:19,080 --> 00:05:21,320
Speaker 3: would pick up the chess piece, move it back as

95
00:05:21,400 --> 00:05:23,760
Speaker 3: if to say like, no, no, Napoleon, let' see what

96
00:05:23,800 --> 00:05:26,120
Speaker 3: you're doing. And then if the person attempted to move

97
00:05:26,120 --> 00:05:28,039
Speaker 3: it again, I don't know how many times on you

98
00:05:28,080 --> 00:05:32,039
Speaker 3: two or three times, eventually it would just go ah

99
00:05:32,080 --> 00:05:34,800
Speaker 3: and wipe his hand across the board and knock off

100
00:05:34,800 --> 00:05:37,839
Speaker 3: all the pieces. Game over, right, Which is pretty great. Yeah,

101
00:05:37,880 --> 00:05:39,520
Speaker 3: it's a nice little feature.

102
00:05:39,440 --> 00:05:43,240
Speaker 1: Yeah it is. But it even showed even more that

103
00:05:43,320 --> 00:05:47,400
Speaker 1: this thing was thinking for itself. Yeah, that's the key here, right. Sure,

104
00:05:47,520 --> 00:05:50,760
Speaker 1: chess had been for a very long time viewed as

105
00:05:51,760 --> 00:05:55,400
Speaker 1: only something that a human would be capable of, because

106
00:05:55,440 --> 00:05:58,000
Speaker 1: it took a human intellect, and there was actually a

107
00:05:58,040 --> 00:06:03,760
Speaker 1: guy in English engineer, I think he was a mechanical engineer.

108
00:06:03,960 --> 00:06:08,240
Speaker 1: His name was Robert Willis. He said that chess was

109
00:06:08,279 --> 00:06:12,159
Speaker 1: in the province of intellect alone. So the idea that

110
00:06:12,200 --> 00:06:16,240
Speaker 1: there was this automaton playing chess blew people away. But

111
00:06:16,480 --> 00:06:19,440
Speaker 1: again people figured out, like, okay, there's something going on here.

112
00:06:19,680 --> 00:06:23,960
Speaker 1: We think that von Kempelin is controlling this thing remotely somehow,

113
00:06:24,279 --> 00:06:27,279
Speaker 1: maybe using magnets or whatever. Other people hit upon the

114
00:06:27,320 --> 00:06:30,960
Speaker 1: idea that there was a small person inside the cabinet

115
00:06:31,240 --> 00:06:34,480
Speaker 1: who would hide when the cab when the workings were shown,

116
00:06:34,520 --> 00:06:37,120
Speaker 1: when the cabinet was open to show the workings, and

117
00:06:37,160 --> 00:06:39,320
Speaker 1: then when the cabinet was closed again and the mechanical

118
00:06:39,320 --> 00:06:42,000
Speaker 1: turk started playing, the person that crawled back out and

119
00:06:42,160 --> 00:06:44,800
Speaker 1: was actually controlling it. This seems to be the case

120
00:06:45,080 --> 00:06:47,960
Speaker 1: that there was a person controlling it, but the idea

121
00:06:48,120 --> 00:06:53,120
Speaker 1: that it was it was a machine that could think

122
00:06:53,520 --> 00:06:57,839
Speaker 1: and beat humans in chess had like kind of unsettling implications.

123
00:06:57,960 --> 00:07:04,120
Speaker 3: Yeah, this author Philip thick Ness, great name, British author

124
00:07:04,720 --> 00:07:09,359
Speaker 3: for sure, Philip Thickness. Yeah, he said, and you know,

125
00:07:09,400 --> 00:07:12,720
Speaker 3: people like you said, all those more complicated explanations. In

126
00:07:12,720 --> 00:07:16,200
Speaker 3: this article you sent, Astuteley points out that he followed

127
00:07:16,200 --> 00:07:19,040
Speaker 3: Occam's razor and basically said, he's got a little kid

128
00:07:19,040 --> 00:07:21,640
Speaker 3: in there. He's got a little a little Bobby Fisher

129
00:07:21,640 --> 00:07:25,040
Speaker 3: in there that's really good at chess, and that's what's

130
00:07:25,040 --> 00:07:28,600
Speaker 3: going on. And other people speculated that other, you know,

131
00:07:28,720 --> 00:07:31,600
Speaker 3: little people might be in there, just adults who would

132
00:07:31,600 --> 00:07:34,040
Speaker 3: fit in there. But then you know, there's the explanation

133
00:07:34,080 --> 00:07:36,640
Speaker 3: that he would open it up and shine a candle

134
00:07:36,640 --> 00:07:41,160
Speaker 3: around and say, you know, nothing to see here everyone,

135
00:07:42,160 --> 00:07:46,040
Speaker 3: So what should we reveal the real deal?

136
00:07:46,520 --> 00:07:48,480
Speaker 1: Sure, I think I did already.

137
00:07:49,000 --> 00:07:51,160
Speaker 3: Well I don't think you spelled it out as.

138
00:07:51,160 --> 00:07:52,160
Speaker 1: Oh, we'll spell it out.

139
00:07:52,320 --> 00:07:54,920
Speaker 3: There was a little person in there, Yeah, not just

140
00:07:54,960 --> 00:07:58,000
Speaker 3: one little person, but they would travel around and recruit people.

141
00:07:58,040 --> 00:08:00,200
Speaker 3: I guess people would get tired of being in there.

142
00:08:00,320 --> 00:08:02,480
Speaker 1: Or they'd forget about them and they'd starve and have

143
00:08:02,560 --> 00:08:03,440
Speaker 1: to replace them.

144
00:08:03,640 --> 00:08:05,680
Speaker 3: But it really was a trick. There was a little

145
00:08:05,720 --> 00:08:08,120
Speaker 3: person in there. They did the same thing as like

146
00:08:08,160 --> 00:08:10,600
Speaker 3: the magic acts, you know, when they saw a person

147
00:08:10,640 --> 00:08:13,880
Speaker 3: in half. It's that the lady just gets into a

148
00:08:13,880 --> 00:08:16,520
Speaker 3: tiny little ball right in one section of that box.

149
00:08:16,680 --> 00:08:18,640
Speaker 1: But my thing is this, like, this is not a

150
00:08:18,720 --> 00:08:20,880
Speaker 1: satisfying explanation to me, Chuck.

151
00:08:20,760 --> 00:08:21,560
Speaker 3: I think it's great.

152
00:08:21,840 --> 00:08:25,080
Speaker 1: How did the person keep up with the board above?

153
00:08:25,800 --> 00:08:28,240
Speaker 3: Well, I mean some I don't know if they ever

154
00:08:28,360 --> 00:08:29,880
Speaker 3: proved exactly how it was going.

155
00:08:30,040 --> 00:08:30,680
Speaker 1: That's what I'm saying.

156
00:08:31,080 --> 00:08:34,960
Speaker 3: Oh, okay, whether or not I think they that the

157
00:08:35,080 --> 00:08:39,200
Speaker 3: zoltar or I'm sorry, the turk was just hollowed out

158
00:08:39,320 --> 00:08:42,240
Speaker 3: and you you would just put your arms through the

159
00:08:42,440 --> 00:08:47,520
Speaker 3: arm hole turk, you would become the turk and the

160
00:08:47,600 --> 00:08:48,320
Speaker 3: churk would fuse.

161
00:08:48,520 --> 00:08:50,839
Speaker 1: That's what some people thought. I think that's what Egar

162
00:08:50,880 --> 00:08:52,439
Speaker 1: Allen Poe thought too.

163
00:08:52,160 --> 00:08:54,040
Speaker 3: He Kidney right.

164
00:08:54,800 --> 00:08:57,720
Speaker 1: Other people thought that there was the the the little

165
00:08:57,760 --> 00:09:01,480
Speaker 1: person was underneath the in the cabin operating the trick

166
00:09:01,520 --> 00:09:02,679
Speaker 1: with levers and stuff like that.

167
00:09:02,720 --> 00:09:04,679
Speaker 3: Well, there could have been a mirror or something, you know,

168
00:09:04,880 --> 00:09:06,960
Speaker 3: I guess, like a telescopic mirror.

169
00:09:07,040 --> 00:09:09,000
Speaker 1: That's what's getting me is how would they keep up

170
00:09:09,040 --> 00:09:09,720
Speaker 1: with the game?

171
00:09:10,000 --> 00:09:10,200
Speaker 3: Right?

172
00:09:10,559 --> 00:09:12,480
Speaker 1: You could keep track of the game, but how could

173
00:09:12,480 --> 00:09:14,200
Speaker 1: you see where the other person moved? You would know

174
00:09:14,200 --> 00:09:15,960
Speaker 1: where you moved, but you wouldn't be able to see

175
00:09:15,960 --> 00:09:17,880
Speaker 1: where the other person moved. That's what I don't get.

176
00:09:17,840 --> 00:09:19,280
Speaker 3: Just mirrors smoking mirrors.

177
00:09:19,320 --> 00:09:21,959
Speaker 1: Maybe so, But the point is is it was a fake.

178
00:09:22,000 --> 00:09:24,640
Speaker 1: It was a fraud, but it raise some really big

179
00:09:24,760 --> 00:09:28,800
Speaker 1: questions about the idea of a machine beating a person

180
00:09:28,840 --> 00:09:30,480
Speaker 1: at something like chess.

181
00:09:30,800 --> 00:09:34,880
Speaker 3: Yeah, and it really peaked the mind of one Charles

182
00:09:34,960 --> 00:09:38,719
Speaker 3: Babbage was he was a kid or young at least

183
00:09:38,760 --> 00:09:40,960
Speaker 3: at the time when he saw the Turk in person,

184
00:09:41,920 --> 00:09:44,680
Speaker 3: and a few years afterward he began work on something

185
00:09:44,720 --> 00:09:48,559
Speaker 3: called the Difference Engine, which was a machine that he

186
00:09:48,640 --> 00:09:54,679
Speaker 3: designed to calculate mathematics automatically. So some point to this

187
00:09:54,840 --> 00:09:58,480
Speaker 3: is kind of maybe the beginnings of humans trying to

188
00:09:58,559 --> 00:09:59,240
Speaker 3: create AI.

189
00:10:00,320 --> 00:10:04,439
Speaker 1: Well yeah, with Babbage's differential machine or difference machine.

190
00:10:04,200 --> 00:10:06,720
Speaker 3: Yeah, difference engine. But at the very least, what this

191
00:10:06,920 --> 00:10:10,160
Speaker 3: is is the first that I know of example of

192
00:10:10,720 --> 00:10:14,840
Speaker 3: man versus machine, even though it was really man versus

193
00:10:14,880 --> 00:10:16,640
Speaker 3: man because it was a man in the machine.

194
00:10:16,760 --> 00:10:20,600
Speaker 1: Right. It was a fraud, Yeah, but it sparked that idea,

195
00:10:20,760 --> 00:10:25,040
Speaker 1: It definitely did. And that's something that like chess in particular,

196
00:10:25,120 --> 00:10:28,480
Speaker 1: has always been like this idea of like, if you

197
00:10:28,480 --> 00:10:30,760
Speaker 1: can teach a machine to play chess, you have really

198
00:10:30,800 --> 00:10:35,000
Speaker 1: achieved a milestone. And there's been you know, plenty of programs,

199
00:10:35,960 --> 00:10:39,880
Speaker 1: most notably Deep Blue, which we'll talk about. But there's

200
00:10:40,040 --> 00:10:45,720
Speaker 1: there's been this idea that like part of AI is chess,

201
00:10:45,800 --> 00:10:49,520
Speaker 1: teaching it to play chess. But they, the people who

202
00:10:49,559 --> 00:10:53,400
Speaker 1: develop AI, never set out to make a chess playing AI,

203
00:10:53,960 --> 00:10:56,320
Speaker 1: just to make a machine that can play chess. That's

204
00:10:56,360 --> 00:10:59,400
Speaker 1: not the point. Chess has always been this way to

205
00:10:59,520 --> 00:11:02,440
Speaker 1: demons straight the progress of artificial intelligence.

206
00:11:02,520 --> 00:11:06,400
Speaker 3: Yeah, because it's a complex game that you can't just

207
00:11:06,520 --> 00:11:09,120
Speaker 3: program it, like it almost has to learn.

208
00:11:09,960 --> 00:11:12,600
Speaker 1: Well, it depends on how you come at it at first, right,

209
00:11:12,679 --> 00:11:16,920
Speaker 1: So initially they did try to program it. Okay, there's

210
00:11:17,040 --> 00:11:23,280
Speaker 1: this from basically nineteen fifty to the about the mid

211
00:11:23,640 --> 00:11:26,880
Speaker 1: like about say nineteen fifty to twenty ten, sixty years right,

212
00:11:27,440 --> 00:11:31,080
Speaker 1: That is how they approached AI and chess is you

213
00:11:31,240 --> 00:11:35,320
Speaker 1: figured out how to break chess down and explain it

214
00:11:35,360 --> 00:11:38,840
Speaker 1: to a computer. Now, what if you could, ideally you

215
00:11:38,880 --> 00:11:43,480
Speaker 1: would have this computer or this AI, this artificial intelligence

216
00:11:44,679 --> 00:11:48,559
Speaker 1: be able to think about the outcome of every possible

217
00:11:49,360 --> 00:11:52,320
Speaker 1: or every possible outcome of a move before making it. Right,

218
00:11:52,800 --> 00:11:55,559
Speaker 1: that's just not possible. Still today we don't have computers

219
00:11:55,559 --> 00:11:57,680
Speaker 1: that can do that. Right, So what you have to

220
00:11:57,679 --> 00:12:00,520
Speaker 1: do is figure out how to create shortcuts for the machine,

221
00:12:00,840 --> 00:12:03,760
Speaker 1: give it best practices, that kind of thing. Yeah, And

222
00:12:03,960 --> 00:12:06,600
Speaker 1: that was actually laid out in nineteen fifty by a

223
00:12:06,640 --> 00:12:09,840
Speaker 1: guy named Claude Shannon who's a father of information theory,

224
00:12:10,320 --> 00:12:12,920
Speaker 1: and he wrote a paper with a pretty on the

225
00:12:12,960 --> 00:12:17,160
Speaker 1: nose title called Programming a Computer for Playing Chess. And

226
00:12:17,200 --> 00:12:18,720
Speaker 1: you have to say it like that when you say

227
00:12:18,760 --> 00:12:19,079
Speaker 1: the name.

228
00:12:19,280 --> 00:12:21,520
Speaker 3: Yeah, it's got a question mark at the end, right, But.

229
00:12:21,520 --> 00:12:26,320
Speaker 1: He laid out two big things. One is creating a

230
00:12:26,760 --> 00:12:31,320
Speaker 1: function of the different moves, and then another one is

231
00:12:31,360 --> 00:12:35,200
Speaker 1: called a mini max. And if those were the two

232
00:12:35,240 --> 00:12:39,800
Speaker 1: things that Shannon laid out, and they established about fifty

233
00:12:39,880 --> 00:12:42,840
Speaker 1: or sixty years of development in teaching an AI to

234
00:12:42,840 --> 00:12:43,439
Speaker 1: play chess.

235
00:12:43,840 --> 00:12:47,040
Speaker 3: Yeah. So this evaluation function is just sort of the base,

236
00:12:48,160 --> 00:12:50,240
Speaker 3: the very basis of it all kind of where it starts,

237
00:12:50,280 --> 00:12:52,679
Speaker 3: which is you a kind of give a number to

238
00:12:52,840 --> 00:12:57,120
Speaker 3: create a numerical evaluation based on the state of the

239
00:12:57,120 --> 00:13:01,280
Speaker 3: board at that moment, right, and assign a real number

240
00:13:01,320 --> 00:13:06,000
Speaker 3: evaluation to it. So the highest number that you would

241
00:13:06,520 --> 00:13:11,920
Speaker 3: shoot for is obviously getting checkmate, getting a king and checkmate.

242
00:13:11,720 --> 00:13:14,880
Speaker 1: Right, right, So what you've just done now is by

243
00:13:14,920 --> 00:13:18,200
Speaker 1: assigning a number to a state like the pieces on

244
00:13:18,240 --> 00:13:21,800
Speaker 1: a board. What you what you've done is to say,

245
00:13:22,080 --> 00:13:24,800
Speaker 1: like shoot for this number. Right, the higher the number,

246
00:13:24,880 --> 00:13:26,800
Speaker 1: like you're going to give this aither rule. Now, the

247
00:13:26,880 --> 00:13:30,320
Speaker 1: higher the number, the more desirable that this move that

248
00:13:30,360 --> 00:13:34,240
Speaker 1: could lead to that higher number. Function. Evaluation function is

249
00:13:34,360 --> 00:13:35,320
Speaker 1: what you want to do.

250
00:13:35,440 --> 00:13:38,080
Speaker 3: Right, like capture the night or capture the queen. Capture

251
00:13:38,080 --> 00:13:40,199
Speaker 3: the queen would have a higher evaluation number.

252
00:13:40,040 --> 00:13:43,200
Speaker 1: Right exactly. So that's the function. Then there's another one

253
00:13:43,280 --> 00:13:46,040
Speaker 1: called the mini max. Yeah, this is pretty great where

254
00:13:46,040 --> 00:13:48,760
Speaker 1: you want to minimize the maximum And this is another

255
00:13:48,760 --> 00:13:50,839
Speaker 1: shortcut that they taught computers.

256
00:13:50,440 --> 00:13:52,640
Speaker 3: The maximum loss that is right, Yeah.

257
00:13:52,600 --> 00:13:55,120
Speaker 1: So what they what they taught computers to do is

258
00:13:55,400 --> 00:13:58,560
Speaker 1: so you no computer can look through an entire game

259
00:13:58,640 --> 00:14:03,880
Speaker 1: every possible outcome, but there are computers that can look

260
00:14:04,160 --> 00:14:08,120
Speaker 1: pretty far down the line at every possible outcome. And

261
00:14:09,360 --> 00:14:13,080
Speaker 1: what you can say is, Okay, you want to find

262
00:14:13,160 --> 00:14:16,640
Speaker 1: the evaluation function that is the worst case scenario, the

263
00:14:16,640 --> 00:14:22,200
Speaker 1: maximum loss, and then find the move that will minimize

264
00:14:22,240 --> 00:14:24,120
Speaker 1: the possibility for that outcome.

265
00:14:24,600 --> 00:14:26,760
Speaker 3: Yeah bye, And this is you're only limited by your

266
00:14:26,760 --> 00:14:29,680
Speaker 3: programming power. But by looking not only at the state

267
00:14:29,720 --> 00:14:31,560
Speaker 3: of the board right now, but if I make this

268
00:14:31,680 --> 00:14:35,480
Speaker 3: move and I move the PAWD to this spot, what

269
00:14:35,600 --> 00:14:38,640
Speaker 3: are the next like three moves possibly that could happen

270
00:14:38,680 --> 00:14:41,720
Speaker 3: as a result of this move. And you're only limited,

271
00:14:41,760 --> 00:14:45,200
Speaker 3: like I said, by programming power. So obviously, the more

272
00:14:45,440 --> 00:14:48,000
Speaker 3: juice you have, the more moves ahead that you can

273
00:14:48,000 --> 00:14:48,920
Speaker 3: look exactly.

274
00:14:49,240 --> 00:14:52,240
Speaker 1: And then they just shy away from ones with the

275
00:14:52,280 --> 00:14:55,400
Speaker 1: higher function number exactly or lower function number, depending on

276
00:14:55,440 --> 00:14:59,280
Speaker 1: how you've programmed it. But they're making these decisions based

277
00:14:59,320 --> 00:15:02,280
Speaker 1: on these rules. And then there's other things you can do,

278
00:15:02,360 --> 00:15:07,880
Speaker 1: like little shortcuts to say, if a decision tree leads

279
00:15:07,920 --> 00:15:13,640
Speaker 1: to the other players King being in checkmate, don't even

280
00:15:13,680 --> 00:15:16,920
Speaker 1: think about that move any further, don't evaluate any longer,

281
00:15:17,000 --> 00:15:18,840
Speaker 1: just abandon it because we would never want to make

282
00:15:18,880 --> 00:15:21,200
Speaker 1: that move. So there's all these shortcuts you can do.

283
00:15:21,360 --> 00:15:24,000
Speaker 1: And that's what they did to teach computers. That's what

284
00:15:24,400 --> 00:15:27,440
Speaker 1: Deep Blue did when it beat Gary Kasparov in nineteen

285
00:15:27,480 --> 00:15:30,720
Speaker 1: ninety seven. It was this huge, massive computer that knew

286
00:15:30,840 --> 00:15:35,000
Speaker 1: a lot of chess, a lot about chess. It had

287
00:15:35,040 --> 00:15:38,800
Speaker 1: a lot of rules, a lot of incredibly intricate programming

288
00:15:38,800 --> 00:15:43,080
Speaker 1: that was extremely sharp, and it actually won. It became

289
00:15:43,160 --> 00:15:47,360
Speaker 1: the first computer to beat an actual human chess grand

290
00:15:47,400 --> 00:15:50,680
Speaker 1: master in like regulation match play.

291
00:15:51,000 --> 00:15:54,200
Speaker 3: Yeah. I mean, and I don't think Kasparov gets enough

292
00:15:54,240 --> 00:15:58,000
Speaker 3: credit for being willing to do this, because it was

293
00:15:58,080 --> 00:16:01,000
Speaker 3: a big deal for him to lose. It was in

294
00:16:01,040 --> 00:16:04,800
Speaker 3: this community and the AI community. It sent shock waves,

295
00:16:04,960 --> 00:16:09,080
Speaker 3: and everyone that was alive remembers, even if you didn't

296
00:16:09,080 --> 00:16:12,320
Speaker 3: know anything about either, one remembers Deep Blue being all

297
00:16:12,320 --> 00:16:14,480
Speaker 3: over the news. It was a really big deal. And

298
00:16:14,560 --> 00:16:16,960
Speaker 3: Kasparov put his name on the line and lost.

299
00:16:17,200 --> 00:16:19,960
Speaker 1: Yeah, And I was wondering, Chuck, how how like you

300
00:16:20,000 --> 00:16:21,960
Speaker 1: would get somebody to do that.

301
00:16:22,480 --> 00:16:24,920
Speaker 3: I'm sure the Mountain of catch.

302
00:16:25,080 --> 00:16:26,840
Speaker 1: I guess that would probably be part of it. But

303
00:16:26,880 --> 00:16:29,760
Speaker 1: I mean, shit, I don't know.

304
00:16:29,840 --> 00:16:31,320
Speaker 3: I bet, I bet that's out there. We just I

305
00:16:31,360 --> 00:16:32,160
Speaker 3: just didn't look it up.

306
00:16:32,280 --> 00:16:36,240
Speaker 1: So that's possible. It's also possible that they said, look, man,

307
00:16:36,360 --> 00:16:38,720
Speaker 1: like this is chess we're talking about or whatever. But really,

308
00:16:38,760 --> 00:16:43,080
Speaker 1: what you're doing is helping advance artificial intelligence.

309
00:16:42,560 --> 00:16:46,480
Speaker 3: Right, because we're not really trying ultimately to win chess games.

310
00:16:46,720 --> 00:16:47,800
Speaker 3: We're trying to cure cancer.

311
00:16:47,840 --> 00:16:50,440
Speaker 1: I mean, yeah, we're going to take your title because

312
00:16:50,480 --> 00:16:52,840
Speaker 1: we're going to beat you, or our machine's going to

313
00:16:52,880 --> 00:16:55,320
Speaker 1: beat you. But even still, you're going to be helping

314
00:16:55,400 --> 00:16:58,840
Speaker 1: with cancer. Think of the cancer Casparov. That's probably what

315
00:16:58,840 --> 00:16:59,160
Speaker 1: they said.

316
00:16:59,200 --> 00:17:00,000
Speaker 3: Should we take a break?

317
00:17:00,160 --> 00:17:00,360
Speaker 1: Yeah?

318
00:17:00,840 --> 00:17:03,680
Speaker 3: Uh wait, well should we tease our special guests first?

319
00:17:04,440 --> 00:17:06,280
Speaker 1: Is he okay? I can smell him.

320
00:17:06,280 --> 00:17:08,159
Speaker 3: I don't think we even said we're gonna have a

321
00:17:08,160 --> 00:17:12,480
Speaker 3: special guest later in the episode, Mister Jonathan Strickland of

322
00:17:12,560 --> 00:17:13,040
Speaker 3: Tech Stuff.

323
00:17:13,320 --> 00:17:13,560
Speaker 1: Nice.

324
00:17:13,600 --> 00:17:15,960
Speaker 3: It's just been a long time since, like years since

325
00:17:15,960 --> 00:17:16,720
Speaker 3: we had Stricken.

326
00:17:16,920 --> 00:17:18,960
Speaker 1: The last time we had Stricken was like two thousand

327
00:17:19,000 --> 00:17:21,440
Speaker 1: and nine with the Necronomicon episode.

328
00:17:21,480 --> 00:17:24,359
Speaker 3: What is what? Where's he been besides sitting in between

329
00:17:24,400 --> 00:17:24,880
Speaker 3: us every day?

330
00:17:24,960 --> 00:17:27,119
Speaker 1: It's been a strickling drought, is what it's been.

331
00:17:27,200 --> 00:17:29,160
Speaker 3: Yeah, so Strickland's coming later, but we're going to come

332
00:17:29,160 --> 00:17:31,919
Speaker 3: back after this and talk a little bit more about

333
00:17:32,280 --> 00:17:53,439
Speaker 3: a man versus machine.

334
00:17:55,480 --> 00:17:59,240
Speaker 1: Okay, dude, So what we just described was how AI

335
00:17:59,720 --> 00:18:03,199
Speaker 1: was to play things like chess or to think like

336
00:18:03,280 --> 00:18:05,480
Speaker 1: you take something, you figure out how to break it

337
00:18:05,560 --> 00:18:10,080
Speaker 1: down into little rules and things that a computer can

338
00:18:10,160 --> 00:18:13,000
Speaker 1: think of, right, and then follow these kind of rules

339
00:18:13,040 --> 00:18:15,879
Speaker 1: to make the best decision. That's how it used to be.

340
00:18:16,800 --> 00:18:20,840
Speaker 1: The way that it's done now that everybody's doing now

341
00:18:21,280 --> 00:18:24,400
Speaker 1: is where you are creating a machine that teaches itself.

342
00:18:24,560 --> 00:18:25,760
Speaker 3: Yeah, that's the jam.

343
00:18:25,840 --> 00:18:29,159
Speaker 1: That was the breakthrough. You may have noticed back in

344
00:18:29,200 --> 00:18:34,760
Speaker 1: about ty thirteen twenty fourteen, all of a sudden, things

345
00:18:34,800 --> 00:18:39,520
Speaker 1: like Siri and Alexa got way better at what they

346
00:18:39,560 --> 00:18:44,320
Speaker 1: are doing. They got way less confused. Really, your navigation

347
00:18:44,440 --> 00:18:47,000
Speaker 1: app got a lot better. And the reason why is

348
00:18:47,040 --> 00:18:50,199
Speaker 1: because this type, this new type of AI, this new

349
00:18:50,240 --> 00:18:54,520
Speaker 1: type of machine learning that can teach itself and learn

350
00:18:54,880 --> 00:18:57,760
Speaker 1: on its own, just hit the scene and they just

351
00:18:57,800 --> 00:19:00,360
Speaker 1: started exploding. And one of the things that they were

352
00:19:00,359 --> 00:19:02,159
Speaker 1: first trained on was games.

353
00:19:02,440 --> 00:19:04,720
Speaker 3: Yeah, and it makes sense. And if you thought chess

354
00:19:04,840 --> 00:19:08,800
Speaker 3: was complicated and difficult, when it comes to these new

355
00:19:08,840 --> 00:19:13,600
Speaker 3: AIS that they're teaching to teach themselves game strategy. They said,

356
00:19:13,720 --> 00:19:16,760
Speaker 3: we might as well dive in to the Chinese strategic

357
00:19:16,840 --> 00:19:21,160
Speaker 3: game Go because it has been called the most complex

358
00:19:21,240 --> 00:19:24,800
Speaker 3: game ever devised by humans. Yeah, and this was actually

359
00:19:24,800 --> 00:19:30,280
Speaker 3: that was actually a quote from Demi Hasabi, a neuroscientist

360
00:19:30,480 --> 00:19:35,840
Speaker 3: and the founder of deep Mind, which was deep Mind.

361
00:19:35,840 --> 00:19:38,440
Speaker 3: They were purchased by Google or were they always part

362
00:19:38,440 --> 00:19:38,879
Speaker 3: of Google.

363
00:19:39,040 --> 00:19:41,280
Speaker 1: I don't know if they were a spun off branch

364
00:19:41,359 --> 00:19:43,640
Speaker 1: or where they were purchased, but it's one of Google's

365
00:19:43,720 --> 00:19:45,159
Speaker 1: AI outfits.

366
00:19:45,280 --> 00:19:47,640
Speaker 3: Well, they're one of the teams, yeah, that are designing

367
00:19:47,640 --> 00:19:50,439
Speaker 3: these new programs. And to give you an idea of

368
00:19:50,440 --> 00:19:53,560
Speaker 3: how complex Go is, it deals with a board with

369
00:19:54,040 --> 00:19:59,520
Speaker 3: different stones and there are ten how do you even

370
00:19:59,560 --> 00:19:59,960
Speaker 3: say that?

371
00:20:00,119 --> 00:20:02,720
Speaker 1: Ten to one hundred and seventieth power, So.

372
00:20:02,720 --> 00:20:06,479
Speaker 3: That means one hundred and seventy zeros and take that

373
00:20:06,560 --> 00:20:09,560
Speaker 3: number and that's the number of possible configurations of a

374
00:20:09,560 --> 00:20:10,359
Speaker 3: go board.

375
00:20:10,359 --> 00:20:14,680
Speaker 1: Right, So, like you say, chess is very complex and complicated,

376
00:20:14,720 --> 00:20:17,760
Speaker 1: and it's very difficult to master Go. And I've never

377
00:20:17,800 --> 00:20:21,160
Speaker 1: played Go, of you no, So it's supposedly it's easy

378
00:20:21,200 --> 00:20:21,879
Speaker 1: to learn.

379
00:20:21,960 --> 00:20:24,840
Speaker 3: Right, but very complicated in its simplicity.

380
00:20:24,440 --> 00:20:28,360
Speaker 1: Right, right, exactly, it's extremely difficult to master. And there

381
00:20:28,400 --> 00:20:30,960
Speaker 1: was a guy in the late nineties and I'm guessing

382
00:20:31,080 --> 00:20:36,560
Speaker 1: that he was saying this after Deep Blue beat Caspar

383
00:20:36,560 --> 00:20:40,919
Speaker 1: ofv It was an astrophysicist from Princeton. He said that

384
00:20:41,040 --> 00:20:43,320
Speaker 1: it would probably be one hundred years before a computer

385
00:20:43,400 --> 00:20:45,760
Speaker 1: beats a human a go. To give you an idea

386
00:20:45,760 --> 00:20:50,040
Speaker 1: of just how complex GO is that deeply would just

387
00:20:50,080 --> 00:20:52,600
Speaker 1: be caspar OFV. And this guy's saying it'll still be

388
00:20:52,640 --> 00:20:55,399
Speaker 1: one hundred years before anyone gets beat at GO by

389
00:20:55,400 --> 00:20:55,920
Speaker 1: a computer.

390
00:20:56,119 --> 00:20:58,399
Speaker 3: And he was someone who knew about this stuff, who

391
00:20:58,440 --> 00:21:01,119
Speaker 3: was an astrophysicist. He was just some schmoe at home

392
00:21:01,160 --> 00:21:02,600
Speaker 3: and drunken as reclining.

393
00:21:03,280 --> 00:21:05,840
Speaker 1: Is making asinine predictions.

394
00:21:07,040 --> 00:21:10,360
Speaker 3: So and again we've said this before, but I want

395
00:21:10,359 --> 00:21:14,159
Speaker 3: to reiterate the people that I think Alpha Go is

396
00:21:14,200 --> 00:21:16,600
Speaker 3: the name of this program. The people that created this

397
00:21:16,680 --> 00:21:20,359
Speaker 3: at Deep Mind, they wanted to stress that this is

398
00:21:21,080 --> 00:21:24,240
Speaker 3: a problem solving program. We're just teaching it this game

399
00:21:24,280 --> 00:21:27,400
Speaker 3: at first, just to make it learn and to see

400
00:21:27,400 --> 00:21:29,399
Speaker 3: if it can get good at what it does. But

401
00:21:30,240 --> 00:21:33,000
Speaker 3: they said it is built with the idea that any

402
00:21:33,080 --> 00:21:36,200
Speaker 3: task that has a lot of data that is unstructured

403
00:21:36,200 --> 00:21:38,119
Speaker 3: and you want to find patterns in the data and

404
00:21:38,160 --> 00:21:40,880
Speaker 3: then decide what to do right, And that's kind of

405
00:21:40,960 --> 00:21:43,720
Speaker 3: like what we were talking about. It It crunches down

406
00:21:44,160 --> 00:21:48,560
Speaker 3: all these possible options aka data to decide what move

407
00:21:48,600 --> 00:21:51,000
Speaker 3: should I make right? And you could apply that Ideally,

408
00:21:51,040 --> 00:21:53,800
Speaker 3: they're going to apply this to Alzheimer's and cancer and

409
00:21:53,840 --> 00:21:54,880
Speaker 3: all sorts of things.

410
00:21:54,880 --> 00:21:58,800
Speaker 1: Right, it's general purpose thinking, right, Yeah, and thinking on

411
00:21:58,840 --> 00:22:02,760
Speaker 1: the fly too, and face with novel stuff. So one

412
00:22:02,800 --> 00:22:05,639
Speaker 1: of the reasons why it's good to use games like

413
00:22:05,760 --> 00:22:09,119
Speaker 1: chess or Go or whatever, those are called perfect information

414
00:22:09,280 --> 00:22:13,359
Speaker 1: games where both players or anybody watching has all the

415
00:22:13,359 --> 00:22:17,560
Speaker 1: information that's available on it. There are definite rules or structure.

416
00:22:18,000 --> 00:22:20,919
Speaker 1: It's a good proving ground. But as we'll see, AI

417
00:22:21,040 --> 00:22:24,080
Speaker 1: makers are getting further and further away from those structure

418
00:22:24,160 --> 00:22:28,280
Speaker 1: games as their AI becomes more and more sophisticated, because

419
00:22:28,760 --> 00:22:33,399
Speaker 1: the structure and the limitations aren't necessarily needed anymore, because

420
00:22:33,440 --> 00:22:35,679
Speaker 1: these things are starting to be able to think on

421
00:22:35,720 --> 00:22:40,080
Speaker 1: their own in a very generalized and even creative way.

422
00:22:40,800 --> 00:22:44,399
Speaker 3: Yeah, it's really really interesting. Yeah, the way that they're

423
00:22:44,520 --> 00:22:48,120
Speaker 3: like you said earlier before the break, that we don't

424
00:22:48,160 --> 00:22:51,119
Speaker 3: have computers that can run all the possibilities. So what

425
00:22:51,160 --> 00:22:53,960
Speaker 3: they teach in the case of Alpha Go, this program

426
00:22:54,080 --> 00:22:59,280
Speaker 3: teaches itself by playing itself in these games and Go specifically,

427
00:23:00,080 --> 00:23:02,360
Speaker 3: and the more it plays itself, the more it learns,

428
00:23:02,400 --> 00:23:06,879
Speaker 3: and the more ability it has during a game to

429
00:23:07,000 --> 00:23:10,639
Speaker 3: choose a move by narrowing down possibilities. So instead of like,

430
00:23:11,080 --> 00:23:15,119
Speaker 3: well there are twenty million different variations here. By playing itself,

431
00:23:15,160 --> 00:23:18,080
Speaker 3: it's able to say, well, in this scenario, they're really

432
00:23:18,080 --> 00:23:22,600
Speaker 3: only fifty different moves that I could or should make, right,

433
00:23:22,800 --> 00:23:24,520
Speaker 3: or that's kind of a simplified way to say it.

434
00:23:24,560 --> 00:23:27,160
Speaker 1: But right, No, but it's true. But that's exactly right.

435
00:23:27,240 --> 00:23:29,560
Speaker 1: And what they're doing is basically the same thing that

436
00:23:29,600 --> 00:23:33,440
Speaker 1: a human does. It's going back to its memory banks, Yeah,

437
00:23:33,480 --> 00:23:36,879
Speaker 1: exact experience, huh, and saying well, I've been faced with

438
00:23:36,960 --> 00:23:38,920
Speaker 1: something like this before, and this is what I used

439
00:23:38,960 --> 00:23:41,840
Speaker 1: and it was successful. Forty out of fifty times. I'll

440
00:23:41,880 --> 00:23:44,919
Speaker 1: do this one. This is a pretty reasonable move. Yeah,

441
00:23:45,000 --> 00:23:46,199
Speaker 1: that is what humans do.

442
00:23:46,440 --> 00:23:48,920
Speaker 3: Yeah. Not only I mean, boy, we screwed up the

443
00:23:48,960 --> 00:23:51,000
Speaker 3: chess episode, but I get the idea that when you're

444
00:23:51,440 --> 00:23:53,919
Speaker 3: a chess master, you don't just think what do the

445
00:23:53,960 --> 00:23:55,440
Speaker 3: numbers say and what does the book say?

446
00:23:55,640 --> 00:23:55,880
Speaker 1: Right?

447
00:23:55,920 --> 00:23:59,520
Speaker 3: But man, I did this move that one time and

448
00:23:59,840 --> 00:24:01,800
Speaker 3: it didn't go as the book said.

449
00:24:02,080 --> 00:24:02,320
Speaker 1: Right.

450
00:24:02,400 --> 00:24:04,600
Speaker 3: So that's now factored into my thinking.

451
00:24:04,560 --> 00:24:09,919
Speaker 1: Right, except imagine being able to learn from scratch and

452
00:24:09,960 --> 00:24:13,360
Speaker 1: get to that point in eight days or eight hours. Yeah,

453
00:24:13,400 --> 00:24:16,760
Speaker 1: So that go team the alpha go the first the

454
00:24:16,800 --> 00:24:19,680
Speaker 1: first iteration of alpha go, I think they started working

455
00:24:19,720 --> 00:24:23,560
Speaker 1: on it in twenty fourteen and in twenty sixteen. At

456
00:24:23,600 --> 00:24:28,120
Speaker 1: the end of twenty sixteen, they unleashed it secretly onto

457
00:24:28,760 --> 00:24:32,919
Speaker 1: an Alpha Go website and it started just wiping the

458
00:24:32,920 --> 00:24:36,360
Speaker 1: floor with everybody. Yeah, everybody's like, this thing's pretty good. Oh,

459
00:24:36,359 --> 00:24:39,800
Speaker 1: it's Alpha Go. That was the end of twenty sixteen.

460
00:24:39,960 --> 00:24:43,760
Speaker 3: Okay, so chess had already come and gone. Like, oh,

461
00:24:43,800 --> 00:24:46,600
Speaker 3: by this point, you can download a program that's like

462
00:24:46,800 --> 00:24:47,840
Speaker 3: Deep Blue, right.

463
00:24:47,680 --> 00:24:50,600
Speaker 1: That was That's a great point. Yeah, like today the

464
00:24:50,680 --> 00:24:54,520
Speaker 1: stuff you played chess with on your laptop is even

465
00:24:54,520 --> 00:24:57,000
Speaker 1: more advanced than Deep Blue was in the nineties, and

466
00:24:57,040 --> 00:25:00,600
Speaker 1: it's just on your laptop. But this is so, this

467
00:25:00,640 --> 00:25:02,840
Speaker 1: is Go. This is the end of twenty sixteen. The

468
00:25:02,920 --> 00:25:08,479
Speaker 1: end of twenty seventeen, Alpha Go was replaced with Alpha

469
00:25:08,520 --> 00:25:12,800
Speaker 1: Go zero. It learned what Alpha Go had taken two

470
00:25:12,920 --> 00:25:17,400
Speaker 1: years or three years to learn in forty days by

471
00:25:17,440 --> 00:25:19,400
Speaker 1: teaching itself, and.

472
00:25:19,320 --> 00:25:22,159
Speaker 3: It beat the master. Yeah, and finally in May of

473
00:25:22,680 --> 00:25:28,919
Speaker 3: twenty seventeen, Alpha Go took on Key g the highest

474
00:25:28,960 --> 00:25:31,479
Speaker 3: ranked Go player in the world. Don't know if he

475
00:25:31,880 --> 00:25:32,719
Speaker 3: or she still is.

476
00:25:33,560 --> 00:25:37,679
Speaker 1: No. Lisa A. Doll is the current or was until

477
00:25:37,880 --> 00:25:39,840
Speaker 1: Alpha Alpha Go beat him?

478
00:25:39,880 --> 00:25:42,320
Speaker 3: Oh man? Yeah, did they get knocked off and Alpha

479
00:25:42,320 --> 00:25:45,280
Speaker 3: Go is the champion? Yeah? Like that's that's not fair.

480
00:25:45,920 --> 00:25:49,880
Speaker 1: I if it's match play and the player, the human

481
00:25:49,920 --> 00:25:53,440
Speaker 1: player is accepted a challenge from the computer, I don't

482
00:25:53,480 --> 00:25:56,280
Speaker 1: see why it wouldn't be the world champion.

483
00:25:56,440 --> 00:26:00,560
Speaker 3: Or do they just now say on websites like human champion?

484
00:26:00,680 --> 00:26:03,160
Speaker 3: Maybe in italics? What's like a sneer?

485
00:26:03,440 --> 00:26:05,840
Speaker 1: Right? Maybe? Yeah?

486
00:26:06,119 --> 00:26:06,679
Speaker 3: Interesting?

487
00:26:06,760 --> 00:26:10,200
Speaker 1: What do they call that? Wetwear? Like your brain, your

488
00:26:10,240 --> 00:26:13,520
Speaker 1: neurons and all that. What instead of hardware? It's wetwear?

489
00:26:13,600 --> 00:26:14,480
Speaker 3: Oh? I don't know about that.

490
00:26:14,840 --> 00:26:16,200
Speaker 1: I think that's the term for it.

491
00:26:16,560 --> 00:26:17,360
Speaker 3: What does that mean? Though?

492
00:26:17,520 --> 00:26:21,480
Speaker 1: It means like you you have a substrate, right your intelligence?

493
00:26:21,520 --> 00:26:25,560
Speaker 1: Your intellect is based on your neurons and they're firing

494
00:26:25,640 --> 00:26:28,040
Speaker 1: all that stuff and it's wet and squishy and meat.

495
00:26:28,680 --> 00:26:31,119
Speaker 1: Then there's hardware that you can do the same thing on,

496
00:26:31,200 --> 00:26:35,000
Speaker 1: you can build intelligence on, but it's hardware, it's not wetwear.

497
00:26:35,119 --> 00:26:35,359
Speaker 3: Oh.

498
00:26:35,400 --> 00:26:39,040
Speaker 1: Interesting, so that's probably it. It's the wetwear champion versus

499
00:26:39,080 --> 00:26:43,600
Speaker 1: the hardware champion. But wetwear is italicized with the sneer.

500
00:26:44,400 --> 00:26:46,760
Speaker 3: So where things really got interesting because you were talking

501
00:26:46,800 --> 00:26:51,040
Speaker 3: earlier about what is it with the chest and go?

502
00:26:51,160 --> 00:26:52,399
Speaker 3: What do they called? What kind of games?

503
00:26:53,359 --> 00:26:55,160
Speaker 1: Perfect information games?

504
00:26:55,240 --> 00:26:58,639
Speaker 3: Right? Then you think And my first thought when you

505
00:26:58,680 --> 00:27:01,359
Speaker 3: said that was well, yeah, and then there's there's games

506
00:27:01,400 --> 00:27:04,919
Speaker 3: like poker like Texas Hold Them where there are a

507
00:27:04,960 --> 00:27:07,880
Speaker 3: set of rules. But poker is not about the set

508
00:27:07,880 --> 00:27:10,439
Speaker 3: of rules. It is about sitting down in front of

509
00:27:10,480 --> 00:27:15,480
Speaker 3: whatever five or six people and lying, bluffing and getting

510
00:27:15,480 --> 00:27:18,520
Speaker 3: away with it in your game face being bluff Like

511
00:27:18,800 --> 00:27:25,159
Speaker 3: there's so many human emotions and contextual clues and micro

512
00:27:25,240 --> 00:27:28,960
Speaker 3: expressions and all these things, like surely you could never

513
00:27:29,080 --> 00:27:33,040
Speaker 3: ever teach a machine to win at Texas Holding Poker.

514
00:27:33,119 --> 00:27:35,840
Speaker 1: Yeah, it'll be one hundred years at least before that happens,

515
00:27:35,880 --> 00:27:36,520
Speaker 1: I predict.

516
00:27:37,440 --> 00:27:41,480
Speaker 3: No, they did it, and more than one team has

517
00:27:41,480 --> 00:27:41,800
Speaker 3: done it.

518
00:27:42,040 --> 00:27:45,399
Speaker 1: Yeah. I read there was one from Carnegie Mellon called

519
00:27:46,240 --> 00:27:52,640
Speaker 1: Liberatus AI Go melon Heads, Yeah, go the Thornton Melons.

520
00:27:53,359 --> 00:27:57,480
Speaker 3: Yeah, I mean that's was The University Alberta has one

521
00:27:57,520 --> 00:27:58,280
Speaker 3: called deep Stack.

522
00:27:58,560 --> 00:28:01,160
Speaker 1: That was the one I read about. And it actually

523
00:28:01,760 --> 00:28:05,440
Speaker 1: here's the thing, like if you read the release on it,

524
00:28:05,600 --> 00:28:07,639
Speaker 1: you're like, you don't know how this thing works?

525
00:28:07,680 --> 00:28:08,639
Speaker 3: Do you really?

526
00:28:08,720 --> 00:28:11,040
Speaker 1: Yeah? And I'm pretty sure they don't fully get it,

527
00:28:11,080 --> 00:28:13,120
Speaker 1: because that's one of the problems. I actually talk about

528
00:28:13,119 --> 00:28:15,080
Speaker 1: this in the Existential Risks series.

529
00:28:15,200 --> 00:28:18,399
Speaker 3: That's scary that is to be released right.

530
00:28:18,280 --> 00:28:22,320
Speaker 1: That there is a type of machine learning where the

531
00:28:22,359 --> 00:28:26,480
Speaker 1: machine teaches itself but we don't really understand how it's teaching.

532
00:28:26,680 --> 00:28:28,919
Speaker 1: Probably the scariest one, right, or what it's learning, but

533
00:28:28,960 --> 00:28:31,080
Speaker 1: that's the most prevalent one. That's what a lot of

534
00:28:31,119 --> 00:28:35,200
Speaker 1: this is is like these machines. It's like here's chess,

535
00:28:36,280 --> 00:28:39,480
Speaker 1: go figure it out and they go, okay, got it.

536
00:28:39,880 --> 00:28:41,920
Speaker 1: How'd you do that? Wouldn't you like to know?

537
00:28:42,400 --> 00:28:45,560
Speaker 3: So that's the scariest presentation you will see on AI

538
00:28:45,800 --> 00:28:47,960
Speaker 3: is when someone says, well, how does all this work?

539
00:28:47,960 --> 00:28:50,840
Speaker 1: And they go, but we just know it can be

540
00:28:50,960 --> 00:28:54,040
Speaker 1: to human at poker. But the thing about deep Stack

541
00:28:54,120 --> 00:28:57,520
Speaker 1: at the University of Alberta is that it learned somehow

542
00:28:58,200 --> 00:29:02,000
Speaker 1: some sort of intuition. Yeah, because that's what's required is

543
00:29:02,040 --> 00:29:05,160
Speaker 1: not just the perfect information where you have all the

544
00:29:05,200 --> 00:29:08,600
Speaker 1: information on the board. It's with poker, you don't know

545
00:29:08,640 --> 00:29:10,760
Speaker 1: what the other person's cards are, and you don't know

546
00:29:10,800 --> 00:29:14,880
Speaker 1: if they're lying or bluffing or what they're doing so

547
00:29:14,920 --> 00:29:18,840
Speaker 1: that's an imperfect information game, so that would require intuition,

548
00:29:18,960 --> 00:29:22,400
Speaker 1: and apparently not one, but two different research groups taught

549
00:29:22,480 --> 00:29:24,040
Speaker 1: AI to into it.

550
00:29:24,240 --> 00:29:28,200
Speaker 3: Yeah, Carnegie Mellon came out in January of twenty seventeen

551
00:29:28,760 --> 00:29:32,600
Speaker 3: with its Liberatus AI and they said they spent twenty

552
00:29:32,720 --> 00:29:36,520
Speaker 3: days playing one hundred and twenty thousand hands of Texas

553
00:29:36,520 --> 00:29:42,440
Speaker 3: hold them with four professional poker players and one and

554
00:29:42,520 --> 00:29:45,240
Speaker 3: smoked them. Basically got up to They weren't playing with

555
00:29:45,240 --> 00:29:48,200
Speaker 3: real money, obviously, but they they that would have been great.

556
00:29:48,240 --> 00:29:50,360
Speaker 1: They were playing with skittles like me as a kid.

557
00:29:50,440 --> 00:29:53,760
Speaker 3: Funded their next project, Liberatus was up by one point

558
00:29:53,760 --> 00:29:56,560
Speaker 3: seven million, and one of the quotes from one of

559
00:29:56,600 --> 00:29:59,600
Speaker 3: the poker players that he made to Wired magazine said,

560
00:30:00,000 --> 00:30:01,880
Speaker 3: felt like I was playing against someone who was cheating,

561
00:30:02,320 --> 00:30:04,520
Speaker 3: Like it could see my cards. I'm not accusing it

562
00:30:04,560 --> 00:30:06,160
Speaker 3: of cheating. It was just that good.

563
00:30:06,440 --> 00:30:06,520
Speaker 2: Right.

564
00:30:07,200 --> 00:30:10,680
Speaker 3: So that's a really interesting thing, man, that they could

565
00:30:11,800 --> 00:30:15,920
Speaker 3: teach self teach a program, or a program could teach

566
00:30:15,960 --> 00:30:20,800
Speaker 3: itself intuition. Right, it's creepy. I thought this part was interesting,

567
00:30:20,840 --> 00:30:25,240
Speaker 3: the Atari stuff. This gets pretty fun. Google deep mind.

568
00:30:26,000 --> 00:30:32,920
Speaker 3: Let it's AI wreak havoc on Atari forty nine different

569
00:30:32,960 --> 00:30:35,440
Speaker 3: Atari twenty six hundred games. See, they could figure out

570
00:30:35,440 --> 00:30:39,520
Speaker 3: how to win, and apparently the most difficult one was

571
00:30:39,560 --> 00:30:43,160
Speaker 3: Miss pac Man, which is a tough game. Still man,

572
00:30:43,160 --> 00:30:46,080
Speaker 3: misspac Man. They nailed it. It's still one of the

573
00:30:46,080 --> 00:30:46,680
Speaker 3: great games.

574
00:30:46,840 --> 00:30:51,920
Speaker 1: But their their game, or their Q deep Q network

575
00:30:52,240 --> 00:30:54,239
Speaker 1: algorithm beat it.

576
00:30:54,840 --> 00:30:55,040
Speaker 3: Yeah.

577
00:30:55,080 --> 00:30:59,200
Speaker 1: I think it got the highest score nine nine points,

578
00:30:59,560 --> 00:31:02,200
Speaker 1: and no human or machine has ever achieved that high

579
00:31:02,240 --> 00:31:02,920
Speaker 1: score from what I.

580
00:31:03,000 --> 00:31:05,520
Speaker 3: Understand, amazing. And the way this one does it, the

581
00:31:05,720 --> 00:31:10,600
Speaker 3: hybrid reward architecture that it uses, is really interesting. It

582
00:31:11,040 --> 00:31:14,080
Speaker 3: says here, it generates a top agent that's like a

583
00:31:14,160 --> 00:31:17,360
Speaker 3: senior manager, and then all these other one hundred and

584
00:31:17,400 --> 00:31:20,800
Speaker 3: fifty individual agents. So it's almost like they've devised this

585
00:31:21,640 --> 00:31:26,160
Speaker 3: artificial structural hierarchy of these little worker agents that go

586
00:31:26,240 --> 00:31:30,440
Speaker 3: out and collect I guess data and then move it

587
00:31:30,520 --> 00:31:33,720
Speaker 3: up the chain to this top agent.

588
00:31:34,160 --> 00:31:38,840
Speaker 1: Right, and then this thing says, Okay, you know, I

589
00:31:38,880 --> 00:31:43,760
Speaker 1: think that you're probably right what these agents are probably doing.

590
00:31:43,760 --> 00:31:46,320
Speaker 1: And I don't know this is exactly true, but there's

591
00:31:46,480 --> 00:31:49,280
Speaker 1: there are models out there like this where the agent

592
00:31:49,360 --> 00:31:52,920
Speaker 1: says this is you have a ninety percent chance of

593
00:31:52,960 --> 00:31:56,760
Speaker 1: success at getting this pellet. If we take this action,

594
00:31:57,200 --> 00:31:59,640
Speaker 1: somebody else says, you've got an eighty two percent chance

595
00:31:59,680 --> 00:32:02,120
Speaker 1: of av this ghost if we go this way. And

596
00:32:02,160 --> 00:32:06,440
Speaker 1: then the top agent, the senior manager, can put all

597
00:32:06,440 --> 00:32:08,080
Speaker 1: this stuff together and say, well, if I listen to

598
00:32:08,120 --> 00:32:10,160
Speaker 1: this guy and this guy not only while I evade

599
00:32:10,160 --> 00:32:13,560
Speaker 1: this ghost, I'll go get this pellet. And it's based

600
00:32:13,600 --> 00:32:18,840
Speaker 1: on what confidence level that the lower agents have in

601
00:32:19,120 --> 00:32:23,280
Speaker 1: success in recommending these moves. And then the top agent

602
00:32:23,320 --> 00:32:24,400
Speaker 1: weighs these things.

603
00:32:24,760 --> 00:32:27,080
Speaker 3: Wow, they should give him a little cap.

604
00:32:27,200 --> 00:32:29,840
Speaker 1: But all this is happening like that. Oh yeah, you

605
00:32:29,840 --> 00:32:31,479
Speaker 1: know what I'm saying. This isn't like well, hold on,

606
00:32:31,520 --> 00:32:33,880
Speaker 1: hold on, everybody, what is Harvey? What do you have

607
00:32:33,920 --> 00:32:36,680
Speaker 1: to say? Well, let's get some Chinese in here and

608
00:32:36,720 --> 00:32:39,440
Speaker 1: hash it out, and everybody sits there in order some

609
00:32:39,520 --> 00:32:41,800
Speaker 1: Chinese food, and then you wait for it to come,

610
00:32:41,880 --> 00:32:44,280
Speaker 1: and then you pick up the meeting from that point on.

611
00:32:44,880 --> 00:32:47,920
Speaker 1: And then finally Harvey gives his idea, but he forgot

612
00:32:47,920 --> 00:32:50,000
Speaker 1: what he was talking about, so he just sits down

613
00:32:50,000 --> 00:32:50,920
Speaker 1: and eats his a roll.

614
00:32:51,520 --> 00:32:55,640
Speaker 3: Well, here's a pretty frightening survey. There was a survey

615
00:32:55,640 --> 00:32:58,440
Speaker 3: of more than three hundred and fifty AI researchers, and

616
00:32:58,480 --> 00:33:00,360
Speaker 3: they have the following things to say, And these are

617
00:33:00,360 --> 00:33:03,000
Speaker 3: the pros that are doing this for a living. They

618
00:33:03,040 --> 00:33:06,800
Speaker 3: predicted that within ten years AI will drive better than

619
00:33:06,800 --> 00:33:09,520
Speaker 3: we do. By twenty forty nine they will be able

620
00:33:09,560 --> 00:33:13,200
Speaker 3: to write a best selling novel. AI will generate this

621
00:33:13,320 --> 00:33:17,440
Speaker 3: and by twenty fifty three be better at performing surgery

622
00:33:17,480 --> 00:33:18,280
Speaker 3: than humans are.

623
00:33:19,080 --> 00:33:22,280
Speaker 1: You know. So again, one of the things that about

624
00:33:22,440 --> 00:33:25,240
Speaker 1: the field of artificial intelligence.

625
00:33:25,120 --> 00:33:26,880
Speaker 3: You know a lot about now famous.

626
00:33:27,920 --> 00:33:31,400
Speaker 1: It is famous for making huge predictions that did not

627
00:33:31,520 --> 00:33:35,320
Speaker 1: pan out. Sure, but you've also seen it's also famous

628
00:33:35,360 --> 00:33:39,000
Speaker 1: for beating predictions that you know have been levied against it.

629
00:33:40,320 --> 00:33:43,400
Speaker 1: But there is something in there, Chuck, that stands out

630
00:33:43,400 --> 00:33:45,960
Speaker 1: to me, and that's the idea of an AI writing

631
00:33:46,000 --> 00:33:50,160
Speaker 1: a novel. Like for a very long time I thought, well, yeah, okay,

632
00:33:50,280 --> 00:33:53,080
Speaker 1: you can teach a robot arm to like put a

633
00:33:53,080 --> 00:33:55,960
Speaker 1: car part or something somewhere if you wanted to just

634
00:33:56,000 --> 00:33:59,400
Speaker 1: follow these mechanical things. Or it can use in auition,

635
00:33:59,520 --> 00:34:02,600
Speaker 1: or it can use logic in reason. But to create

636
00:34:02,680 --> 00:34:05,640
Speaker 1: that's different, right. That was like the new frontier. It

637
00:34:05,720 --> 00:34:08,200
Speaker 1: used to be chess and then then was go. The

638
00:34:08,239 --> 00:34:11,719
Speaker 1: next frontier is creativity and they're starting to bang on

639
00:34:11,800 --> 00:34:15,640
Speaker 1: that door big time. There's a game designing AI called

640
00:34:15,680 --> 00:34:19,120
Speaker 1: Angelina out of the University of Foulmouth, which I always

641
00:34:19,160 --> 00:34:21,799
Speaker 1: want to say Foulmouth, Yeah, but we'll just call it

642
00:34:21,840 --> 00:34:26,359
Speaker 1: Foulmouth like it's supposed to. And Angelina actually comes up

643
00:34:26,440 --> 00:34:32,320
Speaker 1: with ideas for new games, not like a different level

644
00:34:32,560 --> 00:34:36,080
Speaker 1: or something like you should put a purple loincloth on

645
00:34:36,120 --> 00:34:38,560
Speaker 1: that player, you know, that'll look kind of cool like

646
00:34:38,920 --> 00:34:42,040
Speaker 1: new games, but whacked out games that humans would never

647
00:34:42,120 --> 00:34:45,560
Speaker 1: think of. One example I saw is in a dungeon

648
00:34:45,760 --> 00:34:49,919
Speaker 1: Battle Royale game, a player controls like ten players at once,

649
00:34:50,239 --> 00:34:52,520
Speaker 1: and some you have to sacrifice to be killed to

650
00:34:52,600 --> 00:34:55,640
Speaker 1: save the others, like the stuff that human wouldn't necessarily

651
00:34:55,640 --> 00:34:57,799
Speaker 1: think of. This AI is coming up with.

652
00:34:58,320 --> 00:35:00,800
Speaker 3: Well, I mean, when you think of creatively, especially something

653
00:35:00,840 --> 00:35:03,080
Speaker 3: like writing a novel or a film, if there are

654
00:35:03,120 --> 00:35:07,000
Speaker 3: only seven stories, I mean, and that's sort of the

655
00:35:07,040 --> 00:35:10,800
Speaker 3: thinking that they're basically every every dramatic story is a

656
00:35:10,880 --> 00:35:12,640
Speaker 3: variation of one of seven things.

657
00:35:12,760 --> 00:35:15,840
Speaker 1: Yeah. So I mean you can look at like AI

658
00:35:16,400 --> 00:35:20,480
Speaker 1: is scary, and in some ways it very much is

659
00:35:20,520 --> 00:35:23,960
Speaker 1: and can be, but there's also like, definitely a level

660
00:35:23,960 --> 00:35:25,920
Speaker 1: of excitement of the whole thing, and the idea that

661
00:35:26,360 --> 00:35:30,239
Speaker 1: there are artificial minds that are coming online or that

662
00:35:30,360 --> 00:35:32,759
Speaker 1: have come online now, that are out there that are

663
00:35:33,000 --> 00:35:36,520
Speaker 1: they'll they'll just naturally by definition, see things differently than

664
00:35:36,560 --> 00:35:38,920
Speaker 1: we do. Yeah, and the idea that they can come

665
00:35:39,000 --> 00:35:40,839
Speaker 1: up with stuff that we've never even thought of there

666
00:35:40,920 --> 00:35:45,560
Speaker 1: is just gonna knock our socks off, hopefully in good ways.

667
00:35:45,320 --> 00:35:48,719
Speaker 1: That's a really cool thing. And so maybe there's just

668
00:35:48,800 --> 00:35:52,360
Speaker 1: seven as far as humans know, but there's an unlimited amount.

669
00:35:52,480 --> 00:35:54,920
Speaker 1: Is if you put computer minds to thinking about these

670
00:35:55,000 --> 00:35:56,719
Speaker 1: kind of things, that's the premise of it.

671
00:35:57,120 --> 00:35:59,800
Speaker 3: Right, so the robot would be like you never thought

672
00:35:59,840 --> 00:36:06,439
Speaker 3: of boy meets girl meets well trilobite. But see even

673
00:36:06,480 --> 00:36:08,439
Speaker 3: that's a variation of right.

674
00:36:09,719 --> 00:36:12,279
Speaker 1: Just imagine something that we've never even thought of.

675
00:36:12,400 --> 00:36:13,920
Speaker 3: Well, do you know how they should do this? If

676
00:36:13,920 --> 00:36:17,719
Speaker 3: they do do that is uh? Is not is just

677
00:36:17,840 --> 00:36:20,520
Speaker 3: release a book and not tell anyone that it was

678
00:36:20,719 --> 00:36:24,359
Speaker 3: written by an AI program because if they do that

679
00:36:24,400 --> 00:36:26,719
Speaker 3: then it's going to be so under scrutiny. Oh yeah,

680
00:36:26,760 --> 00:36:29,600
Speaker 3: they should secretly release this book and then after it's

681
00:36:29,640 --> 00:36:34,440
Speaker 3: a New York Times bestseller, say meet the Whopper, the

682
00:36:34,520 --> 00:36:35,480
Speaker 3: author of this.

683
00:36:36,160 --> 00:36:39,880
Speaker 1: You know his interests are roller skating, playing Tic tac toe,

684
00:36:39,960 --> 00:36:41,680
Speaker 1: and global thermal nuclear war.

685
00:36:42,440 --> 00:36:44,240
Speaker 3: All right, should we take a break and get strickland

686
00:36:44,239 --> 00:36:44,440
Speaker 3: in here.

687
00:36:44,520 --> 00:36:46,680
Speaker 1: Yeah, we're going to end the Strickland drought because it

688
00:36:46,760 --> 00:36:49,040
Speaker 1: is about to rain strickland in this piece.

689
00:36:49,239 --> 00:36:49,920
Speaker 3: You gross.

690
00:37:12,920 --> 00:37:17,080
Speaker 1: Okay, we're back and get this. The scent of strick

691
00:37:17,600 --> 00:37:20,399
Speaker 1: has permeated our place. That's a beautiful scent.

692
00:37:20,520 --> 00:37:25,040
Speaker 3: It smells like a soldering gun and a circuit board

693
00:37:25,360 --> 00:37:27,600
Speaker 3: and feel a lavender in a protein bar.

694
00:37:28,960 --> 00:37:31,319
Speaker 2: That's fair. I was gonna say Draco noir that would

695
00:37:31,320 --> 00:37:31,960
Speaker 2: have been a lie.

696
00:37:32,280 --> 00:37:35,960
Speaker 1: Is that how you said? I always called a drakar dracr.

697
00:37:36,320 --> 00:37:37,160
Speaker 2: That's that's fair.

698
00:37:38,480 --> 00:37:42,640
Speaker 3: I always pronounced it Benetton colors. That was what I wore.

699
00:37:43,280 --> 00:37:44,200
Speaker 1: Oh is that what you wore?

700
00:37:44,320 --> 00:37:46,600
Speaker 3: Yeah? During my what I call the Year of Cologne,

701
00:37:47,880 --> 00:37:48,879
Speaker 3: I had a couple of seven.

702
00:37:51,040 --> 00:37:51,440
Speaker 1: Uh.

703
00:37:51,920 --> 00:37:54,399
Speaker 2: This is scintillating. Why why am I here?

704
00:37:55,600 --> 00:37:58,120
Speaker 1: So we know that you already know because we talked

705
00:37:58,239 --> 00:38:00,480
Speaker 1: via email about this, but we'll tell everybody nobody else.

706
00:38:01,000 --> 00:38:03,319
Speaker 1: We have brought you in here because you are the

707
00:38:03,360 --> 00:38:06,520
Speaker 1: master of tech and we were talking tech today, which

708
00:38:06,560 --> 00:38:08,799
Speaker 1: we've talked about without you before. But frankly, Chuck and

709
00:38:08,840 --> 00:38:11,040
Speaker 1: I and Jerry huddled and we said, this is not

710
00:38:11,120 --> 00:38:14,800
Speaker 1: quite as good with that strict so let's try something different, gotcha.

711
00:38:14,920 --> 00:38:19,799
Speaker 2: And we're talking about games and machine versus man and

712
00:38:20,440 --> 00:38:24,120
Speaker 2: that that whole evolution and how that's gone super crazy

713
00:38:24,120 --> 00:38:24,879
Speaker 2: over the last few years.

714
00:38:24,960 --> 00:38:28,239
Speaker 3: Games without frontiers, as Peter Gabriel would say, yeah, or

715
00:38:28,239 --> 00:38:28,839
Speaker 3: without fear.

716
00:38:29,000 --> 00:38:30,800
Speaker 1: And we've talked, I mean, we've talked a lot about

717
00:38:31,080 --> 00:38:34,880
Speaker 1: the evolution of machine learning and how now it's starting

718
00:38:34,920 --> 00:38:37,680
Speaker 1: to take off like a rocket because they can teach themselves, right.

719
00:38:38,000 --> 00:38:42,040
Speaker 1: But one thing we haven't really talked about are solved games.

720
00:38:42,040 --> 00:38:44,640
Speaker 1: I mean, we talked about chess. Yeah, we talked about

721
00:38:44,680 --> 00:38:48,240
Speaker 1: go right, would those constitute solve games?

722
00:38:48,320 --> 00:38:48,840
Speaker 3: Not really?

723
00:38:49,120 --> 00:38:52,200
Speaker 2: So a solved game is the concept where if you

724
00:38:52,239 --> 00:38:55,760
Speaker 2: were to assume perfect play on either sides of the game,

725
00:38:56,280 --> 00:38:58,560
Speaker 2: you would always know how it was going to end.

726
00:38:58,440 --> 00:39:01,239
Speaker 3: Which we always assume perfect play, right, Yeah, it's kind

727
00:39:01,239 --> 00:39:01,640
Speaker 3: of our.

728
00:39:01,560 --> 00:39:02,479
Speaker 1: Bags, right stuff.

729
00:39:02,480 --> 00:39:05,120
Speaker 2: You should know motto so perfect play just meaning that

730
00:39:05,160 --> 00:39:07,319
Speaker 2: no one ever makes a mistake, so very much the

731
00:39:07,320 --> 00:39:09,440
Speaker 2: way I do my work right stuff, you should know

732
00:39:09,840 --> 00:39:12,920
Speaker 2: exactly So if you were to take a game like

733
00:39:12,960 --> 00:39:15,960
Speaker 2: Tic Tac Toe and you assume perfect play on both sides,

734
00:39:16,280 --> 00:39:17,360
Speaker 2: it is always going to end in.

735
00:39:17,400 --> 00:39:19,800
Speaker 3: A draw, which is what was in war games.

736
00:39:19,920 --> 00:39:22,200
Speaker 2: Yes, right, the only way to win is not to

737
00:39:22,200 --> 00:39:26,600
Speaker 2: play right. Yes, so a game with like a game

738
00:39:26,680 --> 00:39:30,200
Speaker 2: like Connect four, whoever goes first is always going to win,

739
00:39:30,680 --> 00:39:33,200
Speaker 2: assuming perfect play both sides.

740
00:39:33,320 --> 00:39:35,960
Speaker 3: Yes, what, I don't think I've played Connect for it.

741
00:39:36,120 --> 00:39:38,160
Speaker 3: That's where you drop ever a long time. That's wh

742
00:39:38,160 --> 00:39:41,280
Speaker 3: where you drop the little tokens.

743
00:39:41,440 --> 00:39:42,760
Speaker 2: Yeah, like it kind of like Checkers.

744
00:39:42,840 --> 00:39:45,920
Speaker 1: You did an interstitial playing Connect four. Remember I was.

745
00:39:45,880 --> 00:39:47,680
Speaker 3: Faking it though, and you had perfect play, so I

746
00:39:47,719 --> 00:39:48,520
Speaker 3: knew it was useless.

747
00:39:48,880 --> 00:39:51,239
Speaker 1: I was going to say that I'm so humiliated by

748
00:39:51,280 --> 00:39:55,120
Speaker 1: all the Connect four games that I've lost starting even.

749
00:39:55,080 --> 00:39:58,560
Speaker 2: Yeah, but I mean perfect play. That's something that that

750
00:39:59,040 --> 00:40:05,280
Speaker 2: obviously only the best players typically achieve with significantly complex games.

751
00:40:05,280 --> 00:40:07,719
Speaker 2: Obviously the simpler the game, the easier it is to

752
00:40:07,840 --> 00:40:11,200
Speaker 2: play perfectly. Right tic Tac toe, if you know, once

753
00:40:11,239 --> 00:40:13,960
Speaker 2: you've mastered the basics of Tic Tac Toe and the

754
00:40:14,000 --> 00:40:17,080
Speaker 2: other person has, you're never really going to win unless

755
00:40:17,080 --> 00:40:20,240
Speaker 2: someone has just made a silly mistake because they weren't

756
00:40:20,239 --> 00:40:20,840
Speaker 2: paying attention.

757
00:40:21,120 --> 00:40:23,160
Speaker 3: They put a star instead of an xer right.

758
00:40:23,480 --> 00:40:27,480
Speaker 1: Which doesn't count automatically disqualified you. One thing I've found

759
00:40:27,520 --> 00:40:29,759
Speaker 1: that's very enjoyable is playing with little kids who haven't

760
00:40:29,760 --> 00:40:32,080
Speaker 1: figured out that tic tac toe is very easy.

761
00:40:31,880 --> 00:40:34,800
Speaker 3: To yes and smash their face on the board and

762
00:40:34,920 --> 00:40:35,480
Speaker 3: rub it in.

763
00:40:35,600 --> 00:40:35,839
Speaker 1: Yeah.

764
00:40:35,960 --> 00:40:37,960
Speaker 2: I mean the same reason why I like to join

765
00:40:38,000 --> 00:40:40,200
Speaker 2: in on little league games because I can really whale

766
00:40:40,239 --> 00:40:42,799
Speaker 2: that ball out of the park. Yeah, I really missed

767
00:40:42,800 --> 00:40:43,480
Speaker 2: me feel like a man.

768
00:40:43,600 --> 00:40:46,080
Speaker 3: That's the most tech stuffy thing you've ever said. You

769
00:40:46,120 --> 00:40:47,600
Speaker 3: really whaled that ball out of the park.

770
00:40:47,760 --> 00:40:50,320
Speaker 2: Well, to be fair, I did just do a textuff

771
00:40:50,320 --> 00:40:52,800
Speaker 2: episode about the technology behind baseball bats, so it's so

772
00:40:52,960 --> 00:40:53,600
Speaker 2: fresh on mine.

773
00:40:53,680 --> 00:40:53,839
Speaker 1: Nice.

774
00:40:53,960 --> 00:40:54,839
Speaker 3: Yeah one.

775
00:40:54,840 --> 00:40:57,239
Speaker 2: Actually, it's a lot of fun. So there have been

776
00:40:57,280 --> 00:40:59,800
Speaker 2: a lot of games that have been solved, but Checkers

777
00:40:59,800 --> 00:41:02,919
Speaker 2: will one that was recently solved back in recently by

778
00:41:02,960 --> 00:41:07,280
Speaker 2: the early nineties when it was played against a computer

779
00:41:07,440 --> 00:41:11,120
Speaker 2: called Chinook and chi in Ok.

780
00:41:11,440 --> 00:41:13,839
Speaker 1: Yeah, like the Helicopter or the wins that blow through

781
00:41:13,880 --> 00:41:15,200
Speaker 1: Alberta exactly.

782
00:41:15,680 --> 00:41:20,440
Speaker 2: And so there are certain games that are more easily

783
00:41:20,480 --> 00:41:23,319
Speaker 2: solved than others. You do it through an algorithm, but

784
00:41:23,560 --> 00:41:26,440
Speaker 2: other games like chess, are more complicated because you can

785
00:41:26,760 --> 00:41:29,120
Speaker 2: In chess you have multiple moves that you can do

786
00:41:29,360 --> 00:41:31,719
Speaker 2: where you can you can move a piece back the

787
00:41:31,760 --> 00:41:34,319
Speaker 2: way you win, right, it's not you're not committed to

788
00:41:34,360 --> 00:41:37,239
Speaker 2: going a specific direction with certain pieces, like with a

789
00:41:37,360 --> 00:41:39,520
Speaker 2: night you know you can you could go right back

790
00:41:39,560 --> 00:41:41,759
Speaker 2: to where you started on the next move if you

791
00:41:41,840 --> 00:41:45,960
Speaker 2: wanted to, and that creates more complexity. So the more

792
00:41:46,000 --> 00:41:48,840
Speaker 2: complex the game, the more difficult it is to solve.

793
00:41:48,920 --> 00:41:52,319
Speaker 2: And some games are not solvable simply because you'll never

794
00:41:52,520 --> 00:41:56,120
Speaker 2: know what the full state of the game is from

795
00:41:56,280 --> 00:41:59,839
Speaker 2: any given moment. Did you have a chance to talk

796
00:42:00,000 --> 00:42:03,880
Speaker 2: about the difference between perfect knowledge and imperfect knowledge in

797
00:42:03,920 --> 00:42:04,239
Speaker 2: a game.

798
00:42:04,480 --> 00:42:06,319
Speaker 3: Yeah, yeah, we talked about that some. Yeah.

799
00:42:06,400 --> 00:42:10,920
Speaker 2: So computers, obviously they do really well if they understand

800
00:42:11,360 --> 00:42:13,439
Speaker 2: the exact state of the game all the way through,

801
00:42:13,920 --> 00:42:15,359
Speaker 2: if they have perfect knowledge.

802
00:42:15,040 --> 00:42:17,160
Speaker 1: All of the informations there on the board.

803
00:42:17,440 --> 00:42:20,760
Speaker 2: Right and all players can see all information at all times.

804
00:42:20,840 --> 00:42:23,680
Speaker 2: But games like poker, which you guys talked about, obviously

805
00:42:23,760 --> 00:42:27,080
Speaker 2: you have imperfect information. You only know part of the

806
00:42:27,200 --> 00:42:29,960
Speaker 2: state of the game. That's why those games have been

807
00:42:30,040 --> 00:42:34,440
Speaker 2: more difficult, more challenging for computers to get better than

808
00:42:34,520 --> 00:42:37,960
Speaker 2: humans until relatively recently, and there have been two major

809
00:42:38,000 --> 00:42:40,480
Speaker 2: ways of doing that. You either throw more processing power

810
00:42:40,520 --> 00:42:43,320
Speaker 2: at it, like you get a supercomputer, or you create

811
00:42:43,440 --> 00:42:47,320
Speaker 2: neural networks, artificial neural networks, and you start teaching computers

812
00:42:47,360 --> 00:42:50,600
Speaker 2: to quote unquote learn the way people do.

813
00:42:51,480 --> 00:42:53,400
Speaker 1: So we talked about that. Yeah, and one of the

814
00:42:53,400 --> 00:42:56,320
Speaker 1: things that we talked about was how there's this idea

815
00:42:56,360 --> 00:43:01,080
Speaker 1: that the programmers, especially say the people who are making

816
00:43:01,120 --> 00:43:03,600
Speaker 1: programs that are playing poker and or getting good at poker,

817
00:43:03,840 --> 00:43:07,600
Speaker 1: aren't exactly sure how the machines are learning to play

818
00:43:07,640 --> 00:43:11,480
Speaker 1: poker or what they're learning. They're just getting better at poker. Yeah,

819
00:43:11,520 --> 00:43:14,920
Speaker 1: do they know how they're learning poker? They just know

820
00:43:14,960 --> 00:43:17,840
Speaker 1: that they're learning poker and that they're good at it. Now, like,

821
00:43:17,880 --> 00:43:20,840
Speaker 1: where's the intuition? How is that being learned?

822
00:43:20,880 --> 00:43:23,560
Speaker 2: An excellent question. The way it typically has learned, especially

823
00:43:23,600 --> 00:43:26,840
Speaker 2: with artificial neural networks, is that you set up the

824
00:43:26,880 --> 00:43:32,000
Speaker 2: computer to play millions of hands of poker that are

825
00:43:32,719 --> 00:43:37,040
Speaker 2: randomly assigned, so it's truly as random as computers can get.

826
00:43:37,480 --> 00:43:39,839
Speaker 2: That's a whole philosophical discussion that I don't think we're

827
00:43:39,880 --> 00:43:44,360
Speaker 2: ready to go into right now. But you have games

828
00:43:44,400 --> 00:43:48,200
Speaker 2: come up where the computer is playing itself millions upon

829
00:43:48,280 --> 00:43:52,400
Speaker 2: millions of times and learning every single time how the

830
00:43:52,440 --> 00:43:56,960
Speaker 2: statistics play out, how different betting strategies play out. It's

831
00:43:57,040 --> 00:44:01,160
Speaker 2: sort of partitioning its own mind to play against itself.

832
00:44:01,280 --> 00:44:03,960
Speaker 2: And through that process, it's as if you, as a

833
00:44:04,000 --> 00:44:08,080
Speaker 2: human player, were playing thousands of games with your friends

834
00:44:08,080 --> 00:44:10,719
Speaker 2: and you start to figure out, Oh, when I have

835
00:44:10,800 --> 00:44:13,640
Speaker 2: these particular cards and they're in my hand, and let's

836
00:44:13,640 --> 00:44:16,200
Speaker 2: say we're playing Texas, hold them and the community cards

837
00:44:16,239 --> 00:44:20,040
Speaker 2: are are these? Then I know that generally speaking, maybe

838
00:44:20,239 --> 00:44:22,919
Speaker 2: three times out of ten I end up winning. Maybe

839
00:44:22,960 --> 00:44:25,480
Speaker 2: I shouldn't bet. Well, the computer is doing that, but

840
00:44:25,560 --> 00:44:28,880
Speaker 2: on a scale that far dwarfs what any human can do,

841
00:44:29,000 --> 00:44:31,799
Speaker 2: and in a fraction in the amount of time, and

842
00:44:31,880 --> 00:44:35,520
Speaker 2: so it's sort of well, it's intuition in the sense

843
00:44:35,560 --> 00:44:38,160
Speaker 2: of it's just done it so much.

844
00:44:38,239 --> 00:44:44,240
Speaker 3: Right, But does that mean it's completely ignoring micro expressions

845
00:44:44,280 --> 00:44:47,440
Speaker 3: and facial cues, so that doesn't even come into play.

846
00:44:47,840 --> 00:44:49,880
Speaker 1: You should say, Strickland just nodded yet, Yeah I was.

847
00:44:49,920 --> 00:44:53,160
Speaker 2: I was waiting for John Well. I still nod when

848
00:44:53,239 --> 00:44:54,839
Speaker 2: I do a solo show. And I do a lot

849
00:44:54,880 --> 00:44:55,960
Speaker 2: of expressive dance.

850
00:44:56,080 --> 00:44:57,720
Speaker 3: What do you think, Jonathan, I don't know, Jonathan.

851
00:44:59,239 --> 00:45:02,360
Speaker 2: It gets lonely in here, guys. No, but yes, what

852
00:45:02,440 --> 00:45:05,440
Speaker 2: you're saying all the tells, right, Yeah, tells that you

853
00:45:05,440 --> 00:45:08,440
Speaker 2: would use as a human player. The computer does not

854
00:45:08,520 --> 00:45:08,879
Speaker 2: pick up.

855
00:45:08,840 --> 00:45:10,320
Speaker 3: A speypically taking data.

856
00:45:10,440 --> 00:45:12,680
Speaker 2: Yes, typically, what it would do is it would study

857
00:45:12,760 --> 00:45:17,280
Speaker 2: the outcomes of the games from a purely statistical expression,

858
00:45:17,360 --> 00:45:19,759
Speaker 2: so that most of these poker games tend to be

859
00:45:19,840 --> 00:45:22,560
Speaker 2: computer based poker games. So it's not that it's playing

860
00:45:22,719 --> 00:45:25,520
Speaker 2: like it's not like there's a computer that says, pushed

861
00:45:25,600 --> 00:45:28,680
Speaker 2: ten more chips into the table. You know, I tick

862
00:45:29,000 --> 00:45:32,879
Speaker 2: right exactly a little it's a little winky face emoticon,

863
00:45:33,120 --> 00:45:36,279
Speaker 2: like I don't have good cards. No, it's it's all

864
00:45:36,360 --> 00:45:40,040
Speaker 2: usually over, sort of like internet poker, which a lot

865
00:45:40,080 --> 00:45:43,279
Speaker 2: of the people who play professional poker cut their teeth on,

866
00:45:43,719 --> 00:45:46,600
Speaker 2: especially you know, in the the more recent generations of

867
00:45:46,960 --> 00:45:48,120
Speaker 2: professional poker players.

868
00:45:48,200 --> 00:45:50,920
Speaker 3: Kids today, Yeah, they don't know what it's like being

869
00:45:50,920 --> 00:45:52,600
Speaker 3: a smoky saloon like Moneymaker.

870
00:45:53,000 --> 00:45:55,520
Speaker 2: When Moneymaker rose to the top a few years ago,

871
00:45:55,880 --> 00:45:58,239
Speaker 2: more than like a decade ago, now, he had come

872
00:45:58,280 --> 00:46:01,440
Speaker 2: from the world of internet poker, and so he was

873
00:46:01,560 --> 00:46:04,760
Speaker 2: using those same sort of skills in a real world setting.

874
00:46:05,040 --> 00:46:09,040
Speaker 2: But obviously there are subtle things that we humans do

875
00:46:09,120 --> 00:46:11,759
Speaker 2: in our expressions their computers do not pick up on it.

876
00:46:11,840 --> 00:46:14,560
Speaker 2: In fact, that leads us sort of into the realm

877
00:46:14,680 --> 00:46:19,080
Speaker 2: of games where computers don't do as well as humans.

878
00:46:19,480 --> 00:46:21,880
Speaker 3: Yeah, is that list you sent a joke or is

879
00:46:21,920 --> 00:46:23,040
Speaker 3: it real? No, that's real.

880
00:46:23,480 --> 00:46:25,080
Speaker 2: It does seem like it's weird, like one of the

881
00:46:25,080 --> 00:46:28,640
Speaker 2: games on there is pictionary, for example, rag or tag. Yeah,

882
00:46:28,880 --> 00:46:31,960
Speaker 2: but these are some of these are They sound silly,

883
00:46:32,000 --> 00:46:33,759
Speaker 2: But when you start to think about them in terms

884
00:46:33,800 --> 00:46:36,920
Speaker 2: of computation and robotics, you start to realize how incredibly

885
00:46:37,000 --> 00:46:41,320
Speaker 2: complex it is from a technical perspective, but incredibly easy

886
00:46:41,360 --> 00:46:44,719
Speaker 2: it is for your average human being. Okay, so with

887
00:46:44,840 --> 00:46:47,680
Speaker 2: humans a game of tag, once you know the basics,

888
00:46:47,719 --> 00:46:49,480
Speaker 2: it's it's all an instinct.

889
00:46:49,800 --> 00:46:50,400
Speaker 1: You know what to do.

890
00:46:50,480 --> 00:46:52,200
Speaker 2: You run after the person you tried to catch up

891
00:46:52,239 --> 00:46:54,759
Speaker 2: with them and tag them. But you also know push

892
00:46:54,800 --> 00:46:57,680
Speaker 2: them ind as you can. Well, if you're Josh, you

893
00:46:57,680 --> 00:46:59,440
Speaker 2: push them as hard as you can. But most of

894
00:46:59,480 --> 00:47:03,880
Speaker 2: us we tag and we're not trying to cause harm. Robots, however,

895
00:47:04,360 --> 00:47:08,480
Speaker 2: robots not so good on you said, I'm just saying

896
00:47:09,440 --> 00:47:13,560
Speaker 2: Isaac Asimov Isaac Asimov's rules of robotics. Aside, robots are

897
00:47:13,600 --> 00:47:16,000
Speaker 2: not very good at judging how hard they have to

898
00:47:16,080 --> 00:47:18,960
Speaker 2: hit something in order to make contact, right, They're not

899
00:47:19,080 --> 00:47:23,640
Speaker 2: as good at even your bipedal robots that walk around

900
00:47:23,680 --> 00:47:25,520
Speaker 2: like people, even the ones that can run and do

901
00:47:25,560 --> 00:47:26,280
Speaker 2: flips and stuff.

902
00:47:26,280 --> 00:47:27,960
Speaker 3: Have you seen that one the other day that the

903
00:47:28,000 --> 00:47:30,960
Speaker 3: footage of that thing running and jumping, it's really impressive

904
00:47:31,120 --> 00:47:32,560
Speaker 3: and super creepy.

905
00:47:32,680 --> 00:47:36,920
Speaker 2: Yeah, but even so, that's that's a clip of the

906
00:47:36,920 --> 00:47:38,759
Speaker 2: best of If you ever if you ever see the

907
00:47:38,760 --> 00:47:41,359
Speaker 2: clips where they show all the times the robots fallen over.

908
00:47:41,560 --> 00:47:44,200
Speaker 3: Yeah, we're pouring hot coffee in someone's head.

909
00:47:44,320 --> 00:47:47,480
Speaker 1: Yes, but they always play those clip shows, toy.

910
00:47:48,320 --> 00:47:52,120
Speaker 2: Yes, this is true. So DARPA had its big robotics

911
00:47:52,160 --> 00:47:56,000
Speaker 2: challenge a few years ago where they had bipedal robots

912
00:47:56,040 --> 00:48:00,880
Speaker 2: tried to go through a scenario that was simulating Fukushima

913
00:48:01,120 --> 00:48:05,399
Speaker 2: nuclear disaster. So the interesting thing was the robot had

914
00:48:05,400 --> 00:48:07,920
Speaker 2: to complete a series of tasks that would have been

915
00:48:08,000 --> 00:48:11,160
Speaker 2: mundane to humans, things like open up a door and

916
00:48:11,280 --> 00:48:14,480
Speaker 2: walk through it and pick up a power tool and

917
00:48:14,560 --> 00:48:17,759
Speaker 2: use it against a wall, and you can watch the

918
00:48:17,800 --> 00:48:21,439
Speaker 2: footage of some of these robots doing things like being

919
00:48:21,480 --> 00:48:25,560
Speaker 2: unable to open the door because they can't tell if

920
00:48:25,560 --> 00:48:28,279
Speaker 2: they need to pull or push or they open the door,

921
00:48:28,280 --> 00:48:30,959
Speaker 2: but then immediately fall over the threshold of the door.

922
00:48:31,560 --> 00:48:34,560
Speaker 2: And when you see that, you realize, as advanced as

923
00:48:34,640 --> 00:48:37,960
Speaker 2: robotics is, as advanced as machine learning has become, and

924
00:48:38,280 --> 00:48:42,880
Speaker 2: as incredible as our technology has progressed, there are still

925
00:48:42,960 --> 00:48:46,080
Speaker 2: things that are fundamentally simple to your average human right

926
00:48:46,239 --> 00:48:50,000
Speaker 2: that are incredibly complicated from a technical standpoint, Like a.

927
00:48:49,960 --> 00:48:52,439
Speaker 3: Six year old can play jinga better than a robot.

928
00:48:52,280 --> 00:48:54,840
Speaker 1: Right right, right, Okay, But the thing is we're talking

929
00:48:54,960 --> 00:48:57,000
Speaker 1: robots here, and as we go more and more and

930
00:48:57,040 --> 00:49:00,479
Speaker 1: more online and our world becomes more and more web

931
00:49:00,520 --> 00:49:06,000
Speaker 1: based rather than reality based, doesn't the the fact that

932
00:49:06,080 --> 00:49:09,080
Speaker 1: a robot can't walk through a door matter less and less,

933
00:49:09,520 --> 00:49:13,879
Speaker 1: and the idea that that machines are learning intellect and

934
00:49:14,000 --> 00:49:18,600
Speaker 1: the robotivity, and you just blew my mind that that's

935
00:49:18,640 --> 00:49:21,200
Speaker 1: becoming more and more vital and important and something we

936
00:49:21,239 --> 00:49:22,120
Speaker 1: should be paying attention.

937
00:49:22,280 --> 00:49:24,839
Speaker 2: It absolutely is something we should pay attention to. I mean,

938
00:49:24,880 --> 00:49:30,080
Speaker 2: we have robotic stock traders, the trading on thousands of

939
00:49:30,160 --> 00:49:32,920
Speaker 2: trades per second, right fast, so fast that we have

940
00:49:33,040 --> 00:49:37,399
Speaker 2: had stock market booms and crashes that last less than

941
00:49:37,440 --> 00:49:39,040
Speaker 2: a second long due to that.

942
00:49:39,360 --> 00:49:41,680
Speaker 3: So that the robot army that will ultimately defeat us

943
00:49:41,760 --> 00:49:43,800
Speaker 3: is not something from the terminator.

944
00:49:43,840 --> 00:49:47,680
Speaker 1: It's invisible, right, it's online, it will be online, it's.

945
00:49:48,239 --> 00:49:51,200
Speaker 2: It's it's what's determining our retirement.

946
00:49:50,960 --> 00:49:56,040
Speaker 1: Right, Yeah, the global economy or our municipal water supply

947
00:49:56,360 --> 00:49:56,960
Speaker 1: or whatever.

948
00:49:57,160 --> 00:49:57,359
Speaker 3: Yeah.

949
00:49:57,440 --> 00:50:01,239
Speaker 2: No, There's the fascinating thing to me about this is

950
00:50:01,760 --> 00:50:06,359
Speaker 2: not just that we're training machine intelligence to learn and

951
00:50:06,440 --> 00:50:09,560
Speaker 2: to perform at a level better than humans, but that

952
00:50:09,640 --> 00:50:13,719
Speaker 2: we're putting a lot of trust in those devices and

953
00:50:13,880 --> 00:50:19,160
Speaker 2: things that have real incredible impact on our lives, significant

954
00:50:19,200 --> 00:50:21,640
Speaker 2: enough impact where if things were to go south, it

955
00:50:21,640 --> 00:50:24,719
Speaker 2: would be really bad for us, and not in that

956
00:50:24,800 --> 00:50:31,440
Speaker 2: Terminator respect. Terminator is a terrifying dystopian science fiction story.

957
00:50:31,760 --> 00:50:34,240
Speaker 2: But then when you realize what could really happen behind

958
00:50:34,239 --> 00:50:37,160
Speaker 2: the scenes, you think, oh, robots don't have to do

959
00:50:37,200 --> 00:50:39,640
Speaker 2: any physical harm to us to really mess things up.

960
00:50:40,200 --> 00:50:44,560
Speaker 2: So there are certainly some cases for us to be

961
00:50:44,800 --> 00:50:48,680
Speaker 2: very vigilant in the way we deploy this artificial ants

962
00:50:48,960 --> 00:50:54,359
Speaker 2: right from the outset exactly, and too depends not necessarily.

963
00:50:55,120 --> 00:50:59,000
Speaker 2: I think I think it's I don't think it's too late,

964
00:50:59,440 --> 00:51:02,759
Speaker 2: but I think it's getting to that point of no

965
00:51:02,920 --> 00:51:04,959
Speaker 2: return very very quickly.

966
00:51:04,640 --> 00:51:06,480
Speaker 3: By December this year. Yeah.

967
00:51:06,520 --> 00:51:09,040
Speaker 2: Well, if you're if you're someone like if you're someone

968
00:51:09,080 --> 00:51:11,440
Speaker 2: like Elon Musk, you'd say, if we don't do something

969
00:51:11,480 --> 00:51:15,399
Speaker 2: now where we're totally going to plummet off the edge

970
00:51:15,400 --> 00:51:15,840
Speaker 2: of the cliff.

971
00:51:15,880 --> 00:51:18,520
Speaker 1: But now is a window that is rapidly closing.

972
00:51:18,680 --> 00:51:21,239
Speaker 2: Yes, yeah, yeah, the now is the now is a

973
00:51:21,280 --> 00:51:23,800
Speaker 2: time where we've got a deadline. We don't know exactly

974
00:51:23,800 --> 00:51:26,960
Speaker 2: when that deadline is going to be up, but we

975
00:51:27,040 --> 00:51:30,040
Speaker 2: know that it's not getting further out it. We're just

976
00:51:30,080 --> 00:51:33,560
Speaker 2: getting closer to that deadline. So, and a lot of

977
00:51:33,560 --> 00:51:38,360
Speaker 2: this is covered in deep conversations and the artificial intelligence

978
00:51:38,400 --> 00:51:42,640
Speaker 2: and machine learning fields that has been going on for ages,

979
00:51:42,680 --> 00:51:45,000
Speaker 2: to the point where you even have bodies like the

980
00:51:45,000 --> 00:51:49,919
Speaker 2: European Union that have debated on concepts like granting personhood

981
00:51:50,000 --> 00:51:55,000
Speaker 2: to artificial intelligence. So this is a really fascinating and

982
00:51:55,080 --> 00:51:58,720
Speaker 2: deep subject that and the games thing is a great

983
00:51:59,040 --> 00:52:02,960
Speaker 2: entry point into have that conversation. Uh, you know. I'm

984
00:52:03,080 --> 00:52:04,880
Speaker 2: lucky if I can win a game of chess against

985
00:52:04,880 --> 00:52:05,719
Speaker 2: another human being.

986
00:52:05,840 --> 00:52:09,239
Speaker 1: Oh yeah, right, so I can't even describe chess.

987
00:52:09,680 --> 00:52:11,400
Speaker 3: Did I did? My big thing is I do that

988
00:52:11,520 --> 00:52:13,120
Speaker 3: night thing. I call it the night shuffle. I just

989
00:52:13,160 --> 00:52:15,480
Speaker 3: move them back and forth. I just cast.

990
00:52:16,120 --> 00:52:19,120
Speaker 2: If I can castle, then I'm so happy.

991
00:52:20,440 --> 00:52:23,840
Speaker 3: And that's the third tech stuffiest thing. They come in threes.

992
00:52:24,160 --> 00:52:26,239
Speaker 1: Well, Strick, thank you for stopping by.

993
00:52:26,520 --> 00:52:28,319
Speaker 3: Should stick around for listener mail.

994
00:52:28,400 --> 00:52:31,600
Speaker 1: I think you should too. I love to and throw

995
00:52:31,640 --> 00:52:33,160
Speaker 1: out any funny comments that you have.

996
00:52:33,840 --> 00:52:36,080
Speaker 2: I'll throw out comments and then Jerry can decide which

997
00:52:36,080 --> 00:52:36,720
Speaker 2: ones are funny.

998
00:52:36,800 --> 00:52:39,319
Speaker 1: Okay, all right, fair enough, all right, So if you

999
00:52:39,360 --> 00:52:42,600
Speaker 1: want to know more about AI, go listen to tech Stuff.

1000
00:52:42,719 --> 00:52:45,120
Speaker 1: Strict does this every week what days.

1001
00:52:45,520 --> 00:52:48,760
Speaker 2: Monday, Tuesday, Wednesday, Thursday and Friday.

1002
00:52:48,880 --> 00:52:54,360
Speaker 1: Wow, that's amazing, buddy. And wherever you find your podcast yep, okay,

1003
00:52:54,520 --> 00:52:56,279
Speaker 1: And you've been doing it for years, so if you

1004
00:52:56,360 --> 00:52:59,480
Speaker 1: love this, there's a whole big backlog, nine hundred plus episode.

1005
00:52:59,520 --> 00:53:02,680
Speaker 3: You're celebrating your ten year as well, right yep, by.

1006
00:53:02,560 --> 00:53:05,279
Speaker 2: Sure am, I'll be We'll be turning ten and tex

1007
00:53:05,320 --> 00:53:06,800
Speaker 2: stuff on June eleventh.

1008
00:53:07,320 --> 00:53:11,319
Speaker 1: Congratulations, well, since I said happy anniversary, it means it's

1009
00:53:11,360 --> 00:53:12,319
Speaker 1: time for listener mail.

1010
00:53:14,320 --> 00:53:19,279
Speaker 3: Guys, I'm gonna call this Matt Groening and cultural relativism

1011
00:53:19,440 --> 00:53:19,839
Speaker 3: about that?

1012
00:53:20,200 --> 00:53:20,640
Speaker 1: Nice?

1013
00:53:20,960 --> 00:53:23,440
Speaker 3: Hey, guys, love your podcast so much. The massive archive

1014
00:53:23,840 --> 00:53:26,279
Speaker 3: makes for endless learning and entertainment. My favorite part is

1015
00:53:26,320 --> 00:53:29,839
Speaker 3: you were such rad guys, including Strickland, and I could

1016
00:53:29,880 --> 00:53:33,240
Speaker 3: totally imagine how did they know? I can totally imagine

1017
00:53:33,239 --> 00:53:36,600
Speaker 3: myself getting a beer with you two, but without Strickland,

1018
00:53:36,800 --> 00:53:40,000
Speaker 3: your Simpsons episodes were absolutely perfect. I still live in

1019
00:53:40,040 --> 00:53:42,759
Speaker 3: Portland and drove on Flanders and Lovejoy Streets a lot.

1020
00:53:42,840 --> 00:53:45,360
Speaker 1: Wait, is this Matt Groening? Okay?

1021
00:53:45,760 --> 00:53:49,040
Speaker 3: Matt Granning Drew Bart in the sidewalk cement behind Lincoln

1022
00:53:49,080 --> 00:53:52,120
Speaker 3: High School in downtown Portland. You can google that. I

1023
00:53:52,120 --> 00:53:55,080
Speaker 3: would like to offer one interesting observation though, I've noticed

1024
00:53:55,080 --> 00:53:57,000
Speaker 3: that on several episodes you guys have said that you

1025
00:53:57,000 --> 00:54:01,759
Speaker 3: are cultural relativists. Is that pronounce right? Yeah? Yeah? But

1026
00:54:01,800 --> 00:54:04,120
Speaker 3: then in nearly every episode I hear you pass moral

1027
00:54:04,200 --> 00:54:07,200
Speaker 3: judgments on all the messed up stuff that people do,

1028
00:54:07,560 --> 00:54:11,319
Speaker 3: whether it's racism, preak shows, or crematoriums bearing bodies on

1029
00:54:11,360 --> 00:54:14,040
Speaker 3: the sly. You guys are never shy to condemn something

1030
00:54:14,360 --> 00:54:16,960
Speaker 3: that deserves to be condemned. Reminds me of something I

1031
00:54:16,960 --> 00:54:21,279
Speaker 3: read from Yale's sociologist Philip Gorsky, who points out that

1032
00:54:21,320 --> 00:54:26,200
Speaker 3: our own relativism is rarely as radical as our theory requires.

1033
00:54:26,640 --> 00:54:30,279
Speaker 3: We can't be complete relativists in our daily lives. He

1034
00:54:30,320 --> 00:54:33,439
Speaker 3: then gives the example of how academic social scientists, where

1035
00:54:33,480 --> 00:54:38,400
Speaker 3: diehard relativists, get furious and moralistic at the data fudging

1036
00:54:38,480 --> 00:54:42,879
Speaker 3: of other researchers. Anyway, love the show, guys, love tech

1037
00:54:42,920 --> 00:54:46,560
Speaker 3: stuff especially, and will forever be indebted to you for

1038
00:54:46,680 --> 00:54:50,760
Speaker 3: your hilarity and knowledgeability. Cheers Jesse Lusco.

1039
00:54:50,800 --> 00:54:52,080
Speaker 1: Ps go tech stuff.

1040
00:54:52,680 --> 00:54:53,640
Speaker 2: That's sweet.

1041
00:54:53,880 --> 00:54:56,600
Speaker 1: Love that. Yeah, thanks a lot, Jesse. There was an

1042
00:54:56,680 --> 00:54:59,279
Speaker 1: actual episode, and I don't remember which one it was,

1043
00:54:59,320 --> 00:55:02,280
Speaker 1: where we a and in our cultural relativism, do you remember,

1044
00:55:02,760 --> 00:55:05,920
Speaker 1: because we used to just be like, no judgment, no judgment, right,

1045
00:55:06,040 --> 00:55:08,120
Speaker 1: we just can't judge, you know, And then finally we

1046
00:55:08,120 --> 00:55:10,160
Speaker 1: were like, you know what, No, that's not true. We

1047
00:55:10,320 --> 00:55:14,120
Speaker 1: changed our philosophy to include the idea that there are

1048
00:55:14,800 --> 00:55:18,520
Speaker 1: moral absolutes that are universal, although sometimes we are just

1049
00:55:18,600 --> 00:55:22,960
Speaker 1: judge even beyond that. Look at us, Yeah, Well, if

1050
00:55:23,000 --> 00:55:24,640
Speaker 1: you want to get in touch with us, you can

1051
00:55:24,840 --> 00:55:28,280
Speaker 1: send us an email to Stuff podcast at HowStuffWorks dot com.

1052
00:55:28,360 --> 00:55:29,920
Speaker 1: You can send John an email to.

1053
00:55:30,320 --> 00:55:32,480
Speaker 2: Tech stuff at HowStuffWorks dot com.

1054
00:55:32,560 --> 00:55:34,319
Speaker 1: Nice, and then hang out with us at our home

1055
00:55:34,360 --> 00:55:38,359
Speaker 1: on the web. Stuff youshould know dot com and just go.

1056
00:55:38,320 --> 00:55:40,839
Speaker 2: To tech Stuff. Just search it in Google. I come

1057
00:55:40,920 --> 00:55:41,600
Speaker 2: up all the time.

1058
00:55:41,840 --> 00:55:42,440
Speaker 1: Fair enough.

1059
00:55:46,040 --> 00:55:48,920
Speaker 2: Stuff you Should Know is a production of iHeartRadio. For

1060
00:55:49,000 --> 00:55:53,200
Speaker 2: more podcasts my heart Radio, visit the iHeartRadio app, Apple Podcasts,

1061
00:55:53,280 --> 00:55:55,160
Speaker 2: or wherever you listen to your favorite shows.