1
00:00:00,240 --> 00:00:03,480
Speaker 1: Now here's a highlight from Coast to coast am on

2
00:00:03,600 --> 00:00:07,600
Speaker 1: iHeart Radio. So, James, there's actually an advocate for witches

3
00:00:07,680 --> 00:00:11,719
Speaker 1: in the Catholic Church during the Inquisition there was, and

4
00:00:12,080 --> 00:00:14,840
Speaker 1: I find that really fascinating that there seems to be

5
00:00:14,840 --> 00:00:18,760
Speaker 1: a very conflicted relationship between the Catholic Church and witches.

6
00:00:19,920 --> 00:00:23,919
Speaker 1: You know, in Exodus twenty two eighteen there's the famous

7
00:00:23,960 --> 00:00:26,439
Speaker 1: line suffer not a witch to live right, which is

8
00:00:26,480 --> 00:00:29,720
Speaker 1: basically that you should not tolerate witchcraft and witches should

9
00:00:29,720 --> 00:00:32,640
Speaker 1: be put to death. It was that line that gave

10
00:00:32,720 --> 00:00:38,839
Speaker 1: birth to the hammer of witches, this Mallius maleficarum. But

11
00:00:38,960 --> 00:00:43,240
Speaker 1: also in the Old Testament we see King Saul consulting

12
00:00:44,200 --> 00:00:48,080
Speaker 1: the witch of Endoor to find out some information on

13
00:00:48,800 --> 00:00:53,000
Speaker 1: from the prophet who had died Samuel. And so again

14
00:00:53,320 --> 00:00:59,440
Speaker 1: we see which is being in service to a great

15
00:00:59,440 --> 00:01:04,000
Speaker 1: Old test him in king. So I went to Spain

16
00:01:05,080 --> 00:01:10,119
Speaker 1: and I learned about this cult of Saint Columba. Cult

17
00:01:10,280 --> 00:01:17,160
Speaker 1: of Saint Columba Columba Columba. This cult flourished during the

18
00:01:17,200 --> 00:01:21,360
Speaker 1: Middle Ages in Spain, mainly again in Northern Spain area,

19
00:01:21,480 --> 00:01:24,200
Speaker 1: and which was basically considered also a bastion of witches

20
00:01:24,280 --> 00:01:27,920
Speaker 1: that whole area, and according to legend, Columba was a

21
00:01:27,959 --> 00:01:30,880
Speaker 1: witch for the ninth century who went out and met

22
00:01:30,880 --> 00:01:33,520
Speaker 1: the spirit of Christ on the road, and Christ told

23
00:01:33,560 --> 00:01:38,600
Speaker 1: her she could not enter heaven unless she converted to Christianity,

24
00:01:38,640 --> 00:01:42,560
Speaker 1: and so she did, but she remained a witch. She

25
00:01:42,640 --> 00:01:46,400
Speaker 1: was eventually martyred and beheaded for her faith, and because

26
00:01:46,440 --> 00:01:50,120
Speaker 1: of that she became a saint, specifically the patron saint

27
00:01:50,120 --> 00:01:53,240
Speaker 1: of witches to this day, not only during the Middle Ages,

28
00:01:53,280 --> 00:01:56,160
Speaker 1: but even to current day in Spain, she acts as

29
00:01:56,200 --> 00:01:58,760
Speaker 1: a protector for witches, you know, interceding on the behalf

30
00:01:58,760 --> 00:02:01,640
Speaker 1: of good witches while against those who would corrupt such

31
00:02:01,680 --> 00:02:05,480
Speaker 1: craft for evil purposes. Now I spoke to another historian

32
00:02:05,600 --> 00:02:07,920
Speaker 1: trying to find is there any validity to this legend.

33
00:02:08,320 --> 00:02:12,800
Speaker 1: Is there any historical evidence that Saint Colomba existed and

34
00:02:13,000 --> 00:02:16,400
Speaker 1: that this historians said basically that he believed that the

35
00:02:16,800 --> 00:02:18,639
Speaker 1: cult of Saint Colombo was based on the fact that

36
00:02:18,680 --> 00:02:21,359
Speaker 1: there were many women back during the Middle Ages, educated

37
00:02:21,400 --> 00:02:25,160
Speaker 1: women who studied the natural worlds, who sought medicine and herbs.

38
00:02:25,200 --> 00:02:28,320
Speaker 1: They were drawing upon pagan knowledge. Now, some were accused

39
00:02:28,320 --> 00:02:30,880
Speaker 1: of witchcraft. They were poisoned by priests. They're burned at

40
00:02:30,880 --> 00:02:33,480
Speaker 1: the stake. But there are others who because this is

41
00:02:33,520 --> 00:02:36,480
Speaker 1: going on, they saw they saw shelter in nunneries and

42
00:02:36,600 --> 00:02:40,680
Speaker 1: monasteries were like Saint Colomba, they could worship Christ. Ye

43
00:02:40,680 --> 00:02:43,000
Speaker 1: at the same time, they were planning secret gardens in

44
00:02:43,040 --> 00:02:45,800
Speaker 1: these convents and in these nunneries, and they were helping

45
00:02:45,800 --> 00:02:48,160
Speaker 1: to treat the sick and afflicted, and so there was

46
00:02:48,200 --> 00:02:51,960
Speaker 1: basically smudging the line between paganism and Christianity. You know,

47
00:02:52,000 --> 00:02:55,520
Speaker 1: here are women nuns that were you know, planting the

48
00:02:55,639 --> 00:02:59,120
Speaker 1: same sort of secret gardens, these pagan gardens. Yet because

49
00:02:59,160 --> 00:03:02,320
Speaker 1: they were within the wall the convent, they were considered

50
00:03:02,320 --> 00:03:07,400
Speaker 1: to be you know, wise women uh none serving Christ.

51
00:03:07,720 --> 00:03:12,200
Speaker 1: Yet if those same gardens were planted outside those convent

52
00:03:12,280 --> 00:03:16,400
Speaker 1: walls and were being cultivated by just some village woman,

53
00:03:16,480 --> 00:03:19,600
Speaker 1: she was a witch. So they believe that the cult

54
00:03:19,639 --> 00:03:22,080
Speaker 1: of Saint Colombo. They there's this legend of Saint Colomba

55
00:03:22,120 --> 00:03:24,119
Speaker 1: goes back to that there were women. There were these

56
00:03:24,160 --> 00:03:28,440
Speaker 1: women that would normally be considered witches, that sought solace,

57
00:03:28,520 --> 00:03:33,000
Speaker 1: sought safety within within convents and nunneries where they could

58
00:03:33,040 --> 00:03:37,040
Speaker 1: still practice these pagan rights again whilst at the same

59
00:03:37,080 --> 00:03:41,119
Speaker 1: time wearing the cloth of nuns. How many witches were

60
00:03:41,160 --> 00:03:44,560
Speaker 1: there percentage wise, James and that, and then in those

61
00:03:44,680 --> 00:03:48,680
Speaker 1: days well again they attribute the number of deaths to

62
00:03:49,280 --> 00:03:53,040
Speaker 1: uh the Malus malfarcom at a sixty thousand souls during

63
00:03:53,080 --> 00:03:57,200
Speaker 1: those during the fourth and many of them probably not witches.

64
00:03:57,800 --> 00:04:01,920
Speaker 1: They were probably just the wrong place of the time exactly.

65
00:04:02,000 --> 00:04:05,920
Speaker 1: And there's there's a history in the Galacial area where

66
00:04:06,320 --> 00:04:07,880
Speaker 1: I use it. I use this quote at the beginning

67
00:04:07,880 --> 00:04:10,120
Speaker 1: of my book. It says there's an old Galacian proper

68
00:04:10,200 --> 00:04:13,240
Speaker 1: from from northern Spain. I don't believe in witches, but

69
00:04:13,320 --> 00:04:17,360
Speaker 1: they do exist, and that is pretty much the philosophy

70
00:04:17,400 --> 00:04:22,279
Speaker 1: of that area. There are caves in the Pyrenees Mountains

71
00:04:22,279 --> 00:04:25,560
Speaker 1: of northern Spain that to this day are considered to

72
00:04:25,600 --> 00:04:32,800
Speaker 1: be areas where Sabbaths are enacted there to this day,

73
00:04:32,839 --> 00:04:36,599
Speaker 1: people that do believe that whiches are still still in

74
00:04:36,640 --> 00:04:40,520
Speaker 1: those mountains, still produce its still practicing different arts, both

75
00:04:40,560 --> 00:04:43,599
Speaker 1: from from the the good side of the Wickan side

76
00:04:43,600 --> 00:04:48,240
Speaker 1: and the bad side, and that they're still celebrated up there.

77
00:04:48,720 --> 00:04:51,200
Speaker 1: Now tie this into the crucible, and how do we

78
00:04:51,240 --> 00:04:55,520
Speaker 1: get into artificial intelligence, how does that tie in? Well,

79
00:04:55,560 --> 00:05:00,400
Speaker 1: this goes back to the fact that the ps secution

80
00:05:00,440 --> 00:05:06,159
Speaker 1: of which is pretty much was a way of persecuting women,

81
00:05:06,320 --> 00:05:09,919
Speaker 1: especially women that dared question the natural order. And at

82
00:05:09,920 --> 00:05:12,120
Speaker 1: this beginning of this book, we do see a group

83
00:05:12,200 --> 00:05:14,760
Speaker 1: of women that have banded together, a group of women

84
00:05:14,800 --> 00:05:19,520
Speaker 1: scientists in present day that have banded together because again,

85
00:05:19,600 --> 00:05:22,360
Speaker 1: even today, especially in the sciences, there's still a lot

86
00:05:22,360 --> 00:05:26,560
Speaker 1: of persecution of female scientists, a lot of sexism in

87
00:05:26,839 --> 00:05:29,120
Speaker 1: the scientific world when it comes to women that are

88
00:05:29,240 --> 00:05:31,960
Speaker 1: that are trying to do research. There's a lot of

89
00:05:31,960 --> 00:05:33,800
Speaker 1: sort of me two type of issues going on in

90
00:05:33,880 --> 00:05:37,279
Speaker 1: that in that field. The thing, for example, the woman

91
00:05:37,320 --> 00:05:40,039
Speaker 1: that won the Nobel Prize for physics this year, she

92
00:05:40,160 --> 00:05:41,960
Speaker 1: was the first woman to win the Nobel Pride in

93
00:05:42,000 --> 00:05:46,440
Speaker 1: physics in one hundred years, which is brilliant exactly. So

94
00:05:46,680 --> 00:05:49,880
Speaker 1: you know, this point the crux of the story is,

95
00:05:49,880 --> 00:05:51,600
Speaker 1: you know, even though we may not be burning women

96
00:05:51,640 --> 00:05:54,680
Speaker 1: at the stake today, there are still some persecution going on.

97
00:05:55,360 --> 00:05:58,760
Speaker 1: So incrucible, we have a group of women scientists they

98
00:05:58,760 --> 00:06:02,960
Speaker 1: actually even take on mantle by calling themselves Brooks's International brooks.

99
00:06:02,960 --> 00:06:05,200
Speaker 1: This is the Portuguese word for witches. They would rather

100
00:06:05,279 --> 00:06:08,760
Speaker 1: than shying away from that designation, they lean into it

101
00:06:09,880 --> 00:06:13,640
Speaker 1: so they Unfortunately, this occurs in Portugal and Spain, and

102
00:06:13,720 --> 00:06:17,920
Speaker 1: it arises attracts the interest of this the shadowy Crucibellum,

103
00:06:18,440 --> 00:06:20,640
Speaker 1: the Crucible that still might exist, a sort of a

104
00:06:20,839 --> 00:06:25,560
Speaker 1: hard core of the Spanish Inquisition. Mara, this AI researcher,

105
00:06:25,760 --> 00:06:29,440
Speaker 1: is demonstrating her craft, her breakthrough. She has an AI

106
00:06:29,520 --> 00:06:34,760
Speaker 1: research and during that seminar, the Crucible attacks them. They

107
00:06:35,000 --> 00:06:38,760
Speaker 1: murders most of the women. Mara escapes with her technology

108
00:06:38,800 --> 00:06:42,440
Speaker 1: and is on the run. And throughout this book we

109
00:06:42,560 --> 00:06:45,880
Speaker 1: see sort of the evolution of a of Eve, which

110
00:06:45,920 --> 00:06:51,160
Speaker 1: is a Mara's Mara's AI program, and there's scenes from

111
00:06:51,760 --> 00:06:54,719
Speaker 1: in this book from Eve's point of view. From the beginning,

112
00:06:54,760 --> 00:06:56,839
Speaker 1: we see this sort of cold calculating machine and we

113
00:06:56,920 --> 00:07:00,520
Speaker 1: see what potential that could become. And I talked to

114
00:07:00,520 --> 00:07:03,400
Speaker 1: a lot of researchers about this, right, you know, how

115
00:07:03,400 --> 00:07:05,600
Speaker 1: close are wired to this threat. Something we're already there,

116
00:07:05,720 --> 00:07:08,039
Speaker 1: something that's just right, it's going to happen in our lifetime.

117
00:07:08,800 --> 00:07:12,760
Speaker 1: And they're all very concerned about what's going to happen

118
00:07:12,760 --> 00:07:16,080
Speaker 1: when this threshold's crossed, because once we have an artificial

119
00:07:16,080 --> 00:07:19,720
Speaker 1: intelligence that's equivalent to our intelligence, it's probably going to

120
00:07:19,720 --> 00:07:22,440
Speaker 1: become a super intelligence very quickly, and it's going to

121
00:07:22,480 --> 00:07:27,480
Speaker 1: be almost incomprehensible, incomprehensible, and potentially a threat. So I

122
00:07:27,560 --> 00:07:30,680
Speaker 1: pose this question to these AI researchers, is it okay, great,

123
00:07:31,120 --> 00:07:34,000
Speaker 1: if this is happening, What can we do to make

124
00:07:34,040 --> 00:07:36,480
Speaker 1: sure that this AI that makes an appearance in the

125
00:07:36,520 --> 00:07:40,320
Speaker 1: next five to ten years, five to fifteen years from

126
00:07:40,360 --> 00:07:43,760
Speaker 1: becoming them alignant intelligence that's going to share this planet

127
00:07:43,800 --> 00:07:49,640
Speaker 1: with us? And their answer, their answer again was, and

128
00:07:49,760 --> 00:07:52,960
Speaker 1: this was this came from a founder of Skype, that

129
00:07:53,160 --> 00:07:55,880
Speaker 1: the answer and the only answer, We're not gonna be

130
00:07:55,880 --> 00:07:58,600
Speaker 1: able to regulate our way out of this situation. We're

131
00:07:58,640 --> 00:08:01,840
Speaker 1: never going to be able to regulate people, uh from

132
00:08:01,880 --> 00:08:04,040
Speaker 1: stopping this happening. A sale kind of valley has sort

133
00:08:04,040 --> 00:08:09,240
Speaker 1: of a weird philosophy of you know, you know, go

134
00:08:09,440 --> 00:08:15,240
Speaker 1: fast or or or die slow. And every two years, yeah,

135
00:08:15,320 --> 00:08:18,840
Speaker 1: just they they pursuing AI. Every everybody's pursuing AI at

136
00:08:19,000 --> 00:08:22,440
Speaker 1: at a very rapid pace right now. It's basically almost

137
00:08:22,480 --> 00:08:25,640
Speaker 1: like a gold rush because it can be extremely profitable

138
00:08:25,680 --> 00:08:29,400
Speaker 1: and extremely powerful tool. So everybody's after it. That threshold

139
00:08:29,440 --> 00:08:32,160
Speaker 1: will be crossed. But the question is what's going to

140
00:08:32,240 --> 00:08:34,520
Speaker 1: what's that, what's that intelligence, what's that intelligence going to

141
00:08:34,600 --> 00:08:36,760
Speaker 1: look like? And how's that going to act? And so

142
00:08:37,080 --> 00:08:39,640
Speaker 1: there are a small core of AI researchers, including the

143
00:08:39,679 --> 00:08:43,640
Speaker 1: founder of Skype, who believes if the answer is not

144
00:08:43,679 --> 00:08:47,800
Speaker 1: to just go for AI, but to pursue the harder

145
00:08:47,880 --> 00:08:52,160
Speaker 1: road to produce the first friendly AI and AI that sympathetic,

146
00:08:52,200 --> 00:08:58,319
Speaker 1: empathetic to humankind, is that possible? That's that is exactly

147
00:08:58,360 --> 00:09:02,439
Speaker 1: my question, Thanks for asking. Majority of that was Mike exactable.

148
00:09:02,600 --> 00:09:05,240
Speaker 1: How can you do that? How do you know it's

149
00:09:05,240 --> 00:09:08,360
Speaker 1: going to be friendly? Tuesdays on the CW, from the

150
00:09:08,400 --> 00:09:12,120
Speaker 1: executive producer of The Vampire Diaries and The Originals, comes

151
00:09:12,120 --> 00:09:15,439
Speaker 1: the incredible story about making Contact. Get ready for the

152
00:09:15,520 --> 00:09:19,120
Speaker 1: highly anticipated arrival of the new series Roswell, New Mexico.

153
00:09:19,559 --> 00:09:23,760
Speaker 1: When Liz Orteca returns to her sleepy yet legendary hometown

154
00:09:23,800 --> 00:09:27,320
Speaker 1: of Roswell, she's reunited with high school crush Max Evans,

155
00:09:27,320 --> 00:09:31,400
Speaker 1: small town cop with an extraordinary secret, drawn together as

156
00:09:31,400 --> 00:09:35,160
Speaker 1: if by some unseen force of attraction. Max reveals to

157
00:09:35,240 --> 00:09:38,000
Speaker 1: Liz that he's an alien who fell to Earth during

158
00:09:38,040 --> 00:09:41,800
Speaker 1: the infamous UFO Crash of nineteen forty seven from another world.

159
00:09:42,080 --> 00:09:44,400
Speaker 1: He and his brother and sister have lived their entire

160
00:09:44,480 --> 00:09:47,559
Speaker 1: lives and secret, never telling the soul the truth about

161
00:09:47,600 --> 00:09:50,960
Speaker 1: what they are or the mind blowing powers they each possess.

162
00:09:51,360 --> 00:09:53,960
Speaker 1: When Max's family finds out that he has broken their

163
00:09:54,040 --> 00:09:56,920
Speaker 1: vow of secrecy, events are set in motion that will

164
00:09:57,000 --> 00:10:00,280
Speaker 1: rock the entire town of Roswell to its core. Don't

165
00:10:00,320 --> 00:10:04,960
Speaker 1: miss the unbelievable news series Roswell, New Mexico, tuesdays only

166
00:10:05,000 --> 00:10:08,880
Speaker 1: on the CW. And so they again, I am calling

167
00:10:09,040 --> 00:10:12,600
Speaker 1: different answers from different AI researchers about different techniques they

168
00:10:12,600 --> 00:10:14,959
Speaker 1: believe that can help with that. And it comes a

169
00:10:15,000 --> 00:10:17,920
Speaker 1: matter of sort of nature versus nurture. You know that

170
00:10:17,920 --> 00:10:20,520
Speaker 1: that's a debate that goes on how much of who

171
00:10:20,520 --> 00:10:22,720
Speaker 1: we are is from our genetics and how much of

172
00:10:22,760 --> 00:10:25,360
Speaker 1: who we are is how we're raised. And the same

173
00:10:25,400 --> 00:10:27,960
Speaker 1: applies to this AI. That potential is going to come

174
00:10:27,960 --> 00:10:31,720
Speaker 1: onto this onto the world stage. Part of way of

175
00:10:31,760 --> 00:10:35,080
Speaker 1: making it friendly is tied to code, A way of

176
00:10:35,120 --> 00:10:37,440
Speaker 1: coding your way to try to make it friendly. But

177
00:10:37,520 --> 00:10:40,520
Speaker 1: also the nurture side it is, you know how to

178
00:10:40,600 --> 00:10:43,520
Speaker 1: raise this AI, so it's going to be friendly. So

179
00:10:43,760 --> 00:10:47,360
Speaker 1: I called all this information from these different A researchers

180
00:10:47,400 --> 00:10:51,400
Speaker 1: about these steps necessary to produce a friendly AI. And

181
00:10:51,559 --> 00:10:56,440
Speaker 1: that's what I used as my template for showing the

182
00:10:56,440 --> 00:10:58,960
Speaker 1: evolution of Eve in my novel is by using some

183
00:10:59,040 --> 00:11:01,600
Speaker 1: of these these these programs. I'll give you one example.

184
00:11:03,360 --> 00:11:06,200
Speaker 1: The basis of a lot of human emotions. What drives

185
00:11:06,200 --> 00:11:10,960
Speaker 1: our emotions is our hormones. Hormones drive our emotions. So

186
00:11:11,120 --> 00:11:15,240
Speaker 1: rather than trying to control the emotions of an an

187
00:11:15,280 --> 00:11:19,559
Speaker 1: AI from the standpoint of controlling there trying to code

188
00:11:19,559 --> 00:11:25,880
Speaker 1: and emotions, IBM has developed what's called the endocrine endocrine

189
00:11:26,080 --> 00:11:29,840
Speaker 1: mirror program, which is basically a hormonal mirror program. Basically,

190
00:11:29,880 --> 00:11:33,440
Speaker 1: they're an algorithm that will mimic the hormones we have

191
00:11:33,480 --> 00:11:36,160
Speaker 1: in our body. And if we can instill these these

192
00:11:36,160 --> 00:11:40,439
Speaker 1: hormone codes into our AI, theoretically and these hormones then

193
00:11:40,440 --> 00:11:42,640
Speaker 1: will be the what we'll bring about the emotions we

194
00:11:42,720 --> 00:11:46,000
Speaker 1: want to engender in the AI. And so that's from

195
00:11:46,040 --> 00:11:50,600
Speaker 1: the aspect of controlling nature of the AI, controlling the

196
00:11:50,640 --> 00:11:53,320
Speaker 1: code of the AI is by imbuing it with these

197
00:11:53,480 --> 00:11:57,240
Speaker 1: these endocrine or hormonal mirror programs so that hopefully it

198
00:11:57,280 --> 00:12:03,680
Speaker 1: will will develop into something that that's emotionally stable weather side.

199
00:12:03,679 --> 00:12:08,120
Speaker 1: That's that's that's controlling the nature side, controlling the nurture side. Again,

200
00:12:08,720 --> 00:12:10,360
Speaker 1: some of the other answers I was getting from these

201
00:12:10,360 --> 00:12:14,840
Speaker 1: AI researchers were you almost treat an AI like a child.

202
00:12:15,160 --> 00:12:20,040
Speaker 1: Is first shelter it, you know, produce a virtual world

203
00:12:20,040 --> 00:12:23,520
Speaker 1: in which is AI can grow and learn slowly about

204
00:12:23,520 --> 00:12:27,040
Speaker 1: the world at large, versus just thrusting it into the

205
00:12:28,080 --> 00:12:33,120
Speaker 1: chaos that is the Internet today, and is keep it

206
00:12:33,160 --> 00:12:36,240
Speaker 1: somewhat confined like you would a child, and slowly let

207
00:12:36,240 --> 00:12:39,800
Speaker 1: it mature within this confined environment before and layer by

208
00:12:39,920 --> 00:12:43,800
Speaker 1: layer exposed this this AI to the world at large,

209
00:12:44,320 --> 00:12:47,160
Speaker 1: almost like a parent would raise a child. If you

210
00:12:48,240 --> 00:12:50,960
Speaker 1: raise a child in the safe, caring environment, you're almost

211
00:12:51,000 --> 00:12:54,199
Speaker 1: likely going to have a stable child. If you are

212
00:12:54,320 --> 00:13:00,160
Speaker 1: very abusive and callous and isolating with this child, you

213
00:13:00,200 --> 00:13:04,240
Speaker 1: may end up with something sociopathic. So that very dangerous

214
00:13:04,280 --> 00:13:08,000
Speaker 1: novel is combining both nature and nurture, both code and

215
00:13:08,360 --> 00:13:11,760
Speaker 1: upbringing to see this AI come to fruition. Now it's

216
00:13:11,760 --> 00:13:14,920
Speaker 1: only a small percentage of AI researchers that are pursuing this.

217
00:13:15,640 --> 00:13:18,479
Speaker 1: You know, if this book serves anything, it's a cautionary

218
00:13:18,480 --> 00:13:22,480
Speaker 1: tale about maybe we need to have more scientists pursuing

219
00:13:22,559 --> 00:13:28,080
Speaker 1: this avenue because if we do cross that threshold with

220
00:13:28,120 --> 00:13:34,200
Speaker 1: an AI that's that's poorly controlled, poorly constructed, thrown into

221
00:13:34,240 --> 00:13:36,880
Speaker 1: an environment that's confusing to it when it's when it's

222
00:13:36,880 --> 00:13:40,920
Speaker 1: developing self awareness, we're going to probably create an invention

223
00:13:40,960 --> 00:13:43,280
Speaker 1: that's going to be our own end. Is that what

224
00:13:43,360 --> 00:13:46,920
Speaker 1: Stephen Hawking was concerned about the Late fantas exactly what

225
00:13:46,960 --> 00:13:49,560
Speaker 1: it was concerned about, and not just him. You know,

226
00:13:49,679 --> 00:13:53,079
Speaker 1: there was a there was a open letter that was

227
00:13:53,160 --> 00:13:59,080
Speaker 1: signed by a group of sciences, different signatory from Elon

228
00:13:59,160 --> 00:14:04,120
Speaker 1: Musk to even Hawking, to leaders of Google and Apple

229
00:14:04,640 --> 00:14:09,640
Speaker 1: basically calling out that we need to slow down on

230
00:14:09,760 --> 00:14:14,320
Speaker 1: our rapid pace towards AI development. We need to rethink

231
00:14:14,440 --> 00:14:17,160
Speaker 1: this path. We need to We're never going to regulate

232
00:14:17,200 --> 00:14:19,800
Speaker 1: our way out of this, because at some point, some country,

233
00:14:20,880 --> 00:14:23,080
Speaker 1: some secret lab is still going to pursue that because

234
00:14:23,080 --> 00:14:25,560
Speaker 1: it is such a valuable prize. But maybe we need

235
00:14:25,600 --> 00:14:29,920
Speaker 1: to look at creating this avatar creating this friendly AI

236
00:14:30,120 --> 00:14:34,080
Speaker 1: so that if a malignant AI ever arises, that we

237
00:14:34,240 --> 00:14:38,640
Speaker 1: have somebody, this AI that can compete an AI that

238
00:14:38,640 --> 00:14:41,600
Speaker 1: will be our avatar, an AI that's going to help

239
00:14:41,680 --> 00:14:44,480
Speaker 1: us deal with any malignancy that does arise, something that

240
00:14:44,520 --> 00:14:51,680
Speaker 1: will save us. Right, exactly, how can AI run a munk?

241
00:14:52,040 --> 00:14:57,480
Speaker 1: What could happen to it? Well, that is the main concern,

242
00:14:57,800 --> 00:15:00,640
Speaker 1: is that we have very little under standing of what's

243
00:15:00,680 --> 00:15:04,720
Speaker 1: going on. Doesn't really disturbed me. Is I kept hearing

244
00:15:04,720 --> 00:15:08,560
Speaker 1: from AI researchers over and over the term an algorithmic

245
00:15:08,680 --> 00:15:13,520
Speaker 1: black box, And what that basically means is that even

246
00:15:13,640 --> 00:15:16,920
Speaker 1: today we do not know on some of these advanced

247
00:15:16,920 --> 00:15:19,560
Speaker 1: AI programs out there. And there's some powerful programs out

248
00:15:19,560 --> 00:15:23,040
Speaker 1: there we can talk about, like Google's Alpha Go and

249
00:15:23,120 --> 00:15:27,160
Speaker 1: Alpha Go zero, some incredible AI machines out there already,

250
00:15:27,240 --> 00:15:31,360
Speaker 1: and what I've learned, they're already passed that. But what

251
00:15:31,520 --> 00:15:36,040
Speaker 1: even at this point. Most AI researchers, they construct the code,

252
00:15:36,560 --> 00:15:39,840
Speaker 1: they build the machine, they put in data and an

253
00:15:39,840 --> 00:15:45,280
Speaker 1: output occurs, they cannot explain oftentimes how that AI came

254
00:15:45,320 --> 00:15:48,720
Speaker 1: to that conclusion, how it digested that data. That's the

255
00:15:48,720 --> 00:15:51,680
Speaker 1: way that this is their answer that's dangerous if they

256
00:15:51,680 --> 00:15:56,400
Speaker 1: don't know highly. It's so dangerous that DARPA has given

257
00:15:56,480 --> 00:15:59,040
Speaker 1: six point five million dollars to the universe, to Oregon

258
00:15:59,200 --> 00:16:03,840
Speaker 1: specifically to try to get a better handle on how

259
00:16:03,880 --> 00:16:06,960
Speaker 1: AIS think, because they think that is again another way

260
00:16:06,960 --> 00:16:11,880
Speaker 1: of preventing a malignant AI from coming on stage is

261
00:16:12,080 --> 00:16:14,480
Speaker 1: we need to understand how it thinks, because right now

262
00:16:14,560 --> 00:16:18,480
Speaker 1: we do not know in many cases this algorithmic black box,

263
00:16:18,600 --> 00:16:21,280
Speaker 1: you know, we it's something was unknowable at this point,

264
00:16:21,920 --> 00:16:23,960
Speaker 1: and so that's one concern is that we don't know

265
00:16:24,000 --> 00:16:28,080
Speaker 1: how these AIS think at this point. Number two, how

266
00:16:28,120 --> 00:16:33,120
Speaker 1: are you going to confine an AI? There was an

267
00:16:33,120 --> 00:16:36,120
Speaker 1: experiment done about ten years ago. It was called the

268
00:16:36,240 --> 00:16:42,760
Speaker 1: AI box experiment. In this experiment, there was a researcher

269
00:16:42,760 --> 00:16:46,120
Speaker 1: in the Bay Area, an AI researcher, very smart, smart, gentleman.

270
00:16:46,600 --> 00:16:48,960
Speaker 1: He was basically put in a virtual chat room and

271
00:16:49,080 --> 00:16:53,120
Speaker 1: locked in there. There was a cash prize. That's going

272
00:16:53,160 --> 00:16:55,920
Speaker 1: to hell, AI researchers, part of me, that's like going

273
00:16:56,000 --> 00:17:00,520
Speaker 1: to hell. Exactly, it is going to hell. So he

274
00:17:00,600 --> 00:17:04,240
Speaker 1: had a he offered cash prize for other AI researchers

275
00:17:04,720 --> 00:17:08,280
Speaker 1: to challenge him to see if he could talk himself out,

276
00:17:08,280 --> 00:17:11,960
Speaker 1: pretending to be AI himself again a human level of intelligence.

277
00:17:12,560 --> 00:17:15,680
Speaker 1: Could he talk himself out of this locked room, out

278
00:17:15,680 --> 00:17:20,160
Speaker 1: of this locked box? So AI researcher after AI researcher.

279
00:17:20,200 --> 00:17:21,480
Speaker 1: They were going to give it a prize if they

280
00:17:21,480 --> 00:17:26,160
Speaker 1: could keep him locked up everybody failed. Listen to more

281
00:17:26,240 --> 00:17:29,760
Speaker 1: Coast to Coast AM every weeknight at one am Eastern

282
00:17:30,000 --> 00:17:32,439
Speaker 1: and go to Coast to Coast am dot com for

283
00:17:32,560 --> 00:17:32,800
Speaker 1: more