1
00:00:00,080 --> 00:00:02,560
Speaker 1: Thanks for listening to the best of Coast to Coast podcast.

2
00:00:02,720 --> 00:00:04,800
Speaker 1: If you want to hear more than just this highlight

3
00:00:04,800 --> 00:00:07,360
Speaker 1: from the program, become a Coast Insider and you can

4
00:00:07,400 --> 00:00:10,560
Speaker 1: listen to the full episode, plus recent shows covering the

5
00:00:10,600 --> 00:00:14,480
Speaker 1: mysterious death of Kurt Kobain. The possibility that government may

6
00:00:14,480 --> 00:00:17,439
Speaker 1: soon reveal the truth about UFOs and the power of

7
00:00:17,480 --> 00:00:21,360
Speaker 1: witchcraft is told by an actual practicing which start listening

8
00:00:21,360 --> 00:00:23,279
Speaker 1: now by heading over to Coast to Coast a m

9
00:00:23,360 --> 00:00:27,080
Speaker 1: dot com at signing up for Coast Insider. Now here's

10
00:00:27,080 --> 00:00:30,800
Speaker 1: a highlight from Coast to Coast AM on iHeart Radio

11
00:00:31,040 --> 00:00:33,720
Speaker 1: and welcome back our final segment with Bart costco as

12
00:00:33,760 --> 00:00:36,000
Speaker 1: we take your phone calls. He is a professor of

13
00:00:36,000 --> 00:00:40,199
Speaker 1: PhD of d, of course, and he is the doctor.

14
00:00:40,600 --> 00:00:43,559
Speaker 1: He is a professor of electrical engineering and Law at

15
00:00:43,560 --> 00:00:47,800
Speaker 1: the University of Southern California and an award winning pioneer

16
00:00:47,840 --> 00:00:51,400
Speaker 1: and author in the machine learning fields of artificial intelligence

17
00:00:51,400 --> 00:00:55,480
Speaker 1: and neural networks. What do you think comes first, Bart uh,

18
00:00:55,640 --> 00:01:01,880
Speaker 1: the bigger breakthroughs an AI or the neural network technology?

19
00:01:01,800 --> 00:01:04,800
Speaker 1: The pretty much the same these days, George. Neural networks

20
00:01:04,959 --> 00:01:07,640
Speaker 1: used to be considered separate from AI. For example, I

21
00:01:07,760 --> 00:01:12,399
Speaker 1: organized the first I Tripoli International Conference. We call that

22
00:01:12,480 --> 00:01:15,760
Speaker 1: the Big Bang of pomputer neural networks. And in those

23
00:01:15,880 --> 00:01:18,200
Speaker 1: days it was very separate. Neural networks was on the

24
00:01:18,200 --> 00:01:21,800
Speaker 1: electrical engineering side. AI was on the computer science side.

25
00:01:21,840 --> 00:01:25,120
Speaker 1: Today they have fully merged. But there's a problem. What

26
00:01:25,280 --> 00:01:28,880
Speaker 1: neural networks do and do real well as recognized patterns.

27
00:01:28,920 --> 00:01:32,160
Speaker 1: They work in an effect at the subconscious level. But

28
00:01:32,240 --> 00:01:35,880
Speaker 1: what they don't do is reason well. And that's the

29
00:01:36,000 --> 00:01:39,280
Speaker 1: next step in AI where we can go beyond merely

30
00:01:39,319 --> 00:01:42,560
Speaker 1: recognizing patterns like your face and an image or looking

31
00:01:42,880 --> 00:01:47,800
Speaker 1: at a cancer padge at cancer pattern in a display

32
00:01:47,840 --> 00:01:52,160
Speaker 1: on the screen, go from that to intelligent reasoning, way

33
00:01:52,200 --> 00:01:56,960
Speaker 1: beyond what we have today. And it's just the way

34
00:01:57,000 --> 00:02:00,320
Speaker 1: the technology is moving is amazing. What is fuel the

35
00:02:00,440 --> 00:02:05,120
Speaker 1: rapid speed of it. What's happening at the computer level,

36
00:02:05,280 --> 00:02:08,600
Speaker 1: it's the rapid increase in ship density, and that is

37
00:02:08,639 --> 00:02:11,440
Speaker 1: to say, the on off circuits and something called Moore's law.

38
00:02:11,480 --> 00:02:14,280
Speaker 1: It was doubling about every eighteen months or two years.

39
00:02:14,280 --> 00:02:16,480
Speaker 1: That slowed down a little bit, but it's still continuing

40
00:02:16,560 --> 00:02:18,639
Speaker 1: and some version of it will continue for a long time.

41
00:02:19,040 --> 00:02:21,920
Speaker 1: So a lot of what you're seeing today is simply

42
00:02:22,160 --> 00:02:25,760
Speaker 1: very very old algorithm, especially for example, new networks running

43
00:02:25,760 --> 00:02:30,040
Speaker 1: on vastly faster computers. Are there other countries that are

44
00:02:30,120 --> 00:02:35,080
Speaker 1: keeping pace with us in this arena? Absolutely? For example,

45
00:02:35,400 --> 00:02:38,760
Speaker 1: is maybe not at the forefront, but it is certainly trying.

46
00:02:38,800 --> 00:02:43,200
Speaker 1: It has I think a superior educational infrastructure. Russia is

47
00:02:43,240 --> 00:02:48,000
Speaker 1: another company. How about India. I think India's is it

48
00:02:48,080 --> 00:02:51,280
Speaker 1: still has that British system, and they have this wonderful

49
00:02:51,360 --> 00:02:55,400
Speaker 1: system called the Indian Institute of Technology. I T we

50
00:02:55,880 --> 00:02:59,600
Speaker 1: all scramble to higher I I t a PhD students

51
00:03:00,200 --> 00:03:03,160
Speaker 1: and it's it trains well and it has a lot

52
00:03:03,200 --> 00:03:07,120
Speaker 1: of entrepreneurs. Between those two countries George, India and Chinese,

53
00:03:07,120 --> 00:03:10,120
Speaker 1: you've got a third of the planet. And many other

54
00:03:10,160 --> 00:03:14,400
Speaker 1: countries are far ahead Europe, for example, Britain. So it's

55
00:03:14,440 --> 00:03:17,680
Speaker 1: not just a US game anymore. It's worldwide. Part Apple

56
00:03:17,760 --> 00:03:22,040
Speaker 1: says that's going to stop revealing individual sales of its iPhones,

57
00:03:22,080 --> 00:03:25,480
Speaker 1: it's iPads and things like that. They're you know, a

58
00:03:25,560 --> 00:03:28,919
Speaker 1: free technology is beginning to slow down. How much more

59
00:03:29,000 --> 00:03:32,600
Speaker 1: can you keep changing things? My gosh, a lot, an

60
00:03:32,600 --> 00:03:36,640
Speaker 1: awful lot, really, absolutely again, most of the algorithms that

61
00:03:36,720 --> 00:03:40,080
Speaker 1: we have don't scale up. This is another problem. They

62
00:03:40,160 --> 00:03:43,560
Speaker 1: run into what we call the curse of dimensionality. And

63
00:03:43,720 --> 00:03:45,280
Speaker 1: it's one of the one of the few things that

64
00:03:45,400 --> 00:03:47,560
Speaker 1: works and keeps working, is that what I mentioned before

65
00:03:47,600 --> 00:03:51,040
Speaker 1: the break Monte Carlo. It doesn't depend upon the number

66
00:03:51,080 --> 00:03:53,120
Speaker 1: of dimensions. So for example, in finance, if you have

67
00:03:53,160 --> 00:03:56,120
Speaker 1: a portfolio with a few stocks and bonds, it's very

68
00:03:56,120 --> 00:03:59,240
Speaker 1: difficult predicting its future. As you add more stocks and

69
00:03:59,360 --> 00:04:02,520
Speaker 1: bonds to the at the dimensionality increases, it comes much

70
00:04:02,560 --> 00:04:06,400
Speaker 1: more difficult. And without a simulation technique like Monte Carlo

71
00:04:06,520 --> 00:04:09,240
Speaker 1: enhanced by computers, we would be pretty much had a

72
00:04:09,240 --> 00:04:12,760
Speaker 1: loss already. So that's going to continue. But the next

73
00:04:13,520 --> 00:04:16,160
Speaker 1: great leap is when we go from an effect of

74
00:04:16,240 --> 00:04:19,040
Speaker 1: subconscious of AI where we are now to the more

75
00:04:19,040 --> 00:04:21,680
Speaker 1: of the conscious part. And fuzzy logic, by the way,

76
00:04:21,760 --> 00:04:25,080
Speaker 1: is a very big part of that other other than

77
00:04:25,400 --> 00:04:29,480
Speaker 1: of course, um like my with my smartphone, I can text,

78
00:04:29,680 --> 00:04:32,479
Speaker 1: I can call, I can go online, I can look

79
00:04:32,480 --> 00:04:35,279
Speaker 1: at the internet, I can email. What more do I

80
00:04:35,320 --> 00:04:37,760
Speaker 1: want to do? Well, one thing you want to do

81
00:04:37,839 --> 00:04:40,920
Speaker 1: is have image recognition. So when you go shopping, you

82
00:04:41,040 --> 00:04:42,560
Speaker 1: come out of the store and so where did I

83
00:04:42,600 --> 00:04:45,320
Speaker 1: park my car? You've lost your car and you want

84
00:04:45,360 --> 00:04:47,560
Speaker 1: to hold up your smartphone so it can scan the

85
00:04:47,600 --> 00:04:51,159
Speaker 1: parking lot and recognize your car or similar cars. That's

86
00:04:51,200 --> 00:04:55,760
Speaker 1: a very tough computational problem, but it is something that's feasible.

87
00:04:55,920 --> 00:04:58,240
Speaker 1: It's something we can do right now if the computers

88
00:04:58,240 --> 00:04:59,839
Speaker 1: go fast. Well, there's a matter of fact, there's a

89
00:05:00,000 --> 00:05:02,680
Speaker 1: company that has that where you can you get a

90
00:05:02,720 --> 00:05:06,680
Speaker 1: little device from them which you keep in your cigarette

91
00:05:06,760 --> 00:05:09,400
Speaker 1: lighter in your car and then you get the app

92
00:05:09,800 --> 00:05:11,960
Speaker 1: and you turn the app on and it finds your

93
00:05:11,960 --> 00:05:15,200
Speaker 1: car for you. Right. But beyond that, just the ability

94
00:05:15,279 --> 00:05:18,719
Speaker 1: to walk through the shopping mall and to recognize different

95
00:05:18,800 --> 00:05:22,560
Speaker 1: kinds of objects. That's something that's beyond the current capability.

96
00:05:22,560 --> 00:05:25,280
Speaker 1: But it's coming. And what's in your smartphone I think

97
00:05:25,279 --> 00:05:28,400
Speaker 1: will likely be in some form embedded in your retina.

98
00:05:28,680 --> 00:05:31,600
Speaker 1: That's where this is headed, I believe. And look at

99
00:05:31,640 --> 00:05:34,520
Speaker 1: the brains behind this stuff, you know, beyond Facebook and

100
00:05:34,560 --> 00:05:37,599
Speaker 1: what they've done for you know, the social networking, whether

101
00:05:37,640 --> 00:05:41,000
Speaker 1: you like it or not. Um look at the technology

102
00:05:41,040 --> 00:05:45,200
Speaker 1: of Uber for example. Uh, you know, I love our

103
00:05:45,240 --> 00:05:47,599
Speaker 1: cab drivers. We had lots of cab drivers who listen

104
00:05:47,640 --> 00:05:51,919
Speaker 1: to our program. Uh. Yet Uber technology has really hurt

105
00:05:51,960 --> 00:05:56,680
Speaker 1: the cab business uh in more ways than one. But

106
00:05:56,920 --> 00:05:59,960
Speaker 1: I mean, what an amazing technology. Bart. You could literally

107
00:06:00,240 --> 00:06:04,839
Speaker 1: see the car of the driver coming to pick you up,

108
00:06:05,640 --> 00:06:09,560
Speaker 1: going down the road, making a turn, stopping, and then

109
00:06:09,600 --> 00:06:11,960
Speaker 1: you see it coming down with its lights on if

110
00:06:11,960 --> 00:06:13,800
Speaker 1: it's at night, and you go look there it is,

111
00:06:13,880 --> 00:06:15,680
Speaker 1: and you look at your phone and it matches it.

112
00:06:16,440 --> 00:06:19,720
Speaker 1: I mean, my gosh, who came up with that? A

113
00:06:19,720 --> 00:06:22,760
Speaker 1: lot of people didn't. That also depends on GPS. But

114
00:06:22,839 --> 00:06:25,640
Speaker 1: what we really want wanted for decades or smart cars

115
00:06:26,160 --> 00:06:28,440
Speaker 1: that you can just sit in the back and I

116
00:06:28,440 --> 00:06:30,400
Speaker 1: don't want to do that. Well, you don't want to

117
00:06:30,400 --> 00:06:32,760
Speaker 1: do it now, George, but in the future I think

118
00:06:33,200 --> 00:06:35,360
Speaker 1: I think you will want to and I definitely because

119
00:06:35,360 --> 00:06:37,839
Speaker 1: of jobs. I don't want to see a driverless truck.

120
00:06:38,960 --> 00:06:41,320
Speaker 1: I think that's a very long way off. There'll still

121
00:06:41,360 --> 00:06:45,240
Speaker 1: be truck drivers. I think, to the first approximation, this

122
00:06:45,240 --> 00:06:48,520
Speaker 1: will assist them and what they do will drive more efficiently.

123
00:06:48,880 --> 00:06:51,359
Speaker 1: But it takes something like Uber, the smart car and

124
00:06:51,480 --> 00:06:55,760
Speaker 1: many other devices. There is an inevitable automation here and

125
00:06:55,800 --> 00:06:58,560
Speaker 1: so some jobs will inevitably be lost or reduced, and

126
00:06:58,600 --> 00:07:04,200
Speaker 1: many many others will be hated. What's next for smart technology?

127
00:07:04,240 --> 00:07:06,640
Speaker 1: I mean, I didn't think of Uberg. I didn't even

128
00:07:06,640 --> 00:07:09,560
Speaker 1: cross my mind. As a matter of fact, it was

129
00:07:09,600 --> 00:07:12,480
Speaker 1: a couple of years ago. Some people in Los Angeles

130
00:07:13,280 --> 00:07:15,960
Speaker 1: has said, Hey, we're going to take an Uber home

131
00:07:15,960 --> 00:07:18,760
Speaker 1: and I want what's that? I mean, they knew about

132
00:07:18,760 --> 00:07:21,520
Speaker 1: it way before I did, and then it started catching

133
00:07:21,600 --> 00:07:23,960
Speaker 1: on and it turned out to be huge. They even

134
00:07:24,000 --> 00:07:28,200
Speaker 1: started advertising on our program a year ago. There's something else, George,

135
00:07:28,240 --> 00:07:31,960
Speaker 1: that is coming. We have a forerunner and that's what's

136
00:07:32,000 --> 00:07:36,040
Speaker 1: called blockchain. And the great example that is bitcoin. Okay

137
00:07:36,280 --> 00:07:39,320
Speaker 1: came into existence almost exactly ten years ago with this

138
00:07:40,240 --> 00:07:43,440
Speaker 1: anonymous genius that goes by the name of Satashi Nakamoto.

139
00:07:43,480 --> 00:07:44,840
Speaker 1: We we gotta who he is. We can give this

140
00:07:44,880 --> 00:07:47,840
Speaker 1: guy a Nobel Prize in economics. It was probably worth

141
00:07:47,880 --> 00:07:51,400
Speaker 1: a hundred billion dollars, probably worth more than Jeff Bezos

142
00:07:51,440 --> 00:07:54,960
Speaker 1: from Amazon well be but he introduced something or they

143
00:07:55,040 --> 00:07:59,120
Speaker 1: introduced something, whoever Satashi is, that is truly radical in

144
00:07:59,120 --> 00:08:02,160
Speaker 1: the smart world. The idea of a secure ledger. If

145
00:08:02,200 --> 00:08:04,280
Speaker 1: I want to sell my house to you or vice versa.

146
00:08:04,360 --> 00:08:06,560
Speaker 1: You had and I go through an scrow, we have

147
00:08:06,560 --> 00:08:10,400
Speaker 1: a trusted third party. Blockchain gets rid of that. It's

148
00:08:10,440 --> 00:08:14,920
Speaker 1: an open ledger, and yet it's cryptographically secure, and that's

149
00:08:14,960 --> 00:08:19,040
Speaker 1: something secure severely lacking, and not just implants and devices

150
00:08:19,080 --> 00:08:21,320
Speaker 1: like that, but really in all of our communications, can

151
00:08:21,360 --> 00:08:24,440
Speaker 1: you trust it? I think you can trust blockchain because

152
00:08:24,480 --> 00:08:28,520
Speaker 1: in order to outsmart it, you would need vast computational facilities,

153
00:08:29,040 --> 00:08:30,680
Speaker 1: and that's one of the genius things about it. But

154
00:08:30,720 --> 00:08:33,120
Speaker 1: blockchain has many many of their applications. In fact, it's

155
00:08:33,120 --> 00:08:36,080
Speaker 1: one of the ways to boost the security of the

156
00:08:36,120 --> 00:08:38,920
Speaker 1: Internet of things. But no one saw that coming. And

157
00:08:39,120 --> 00:08:41,920
Speaker 1: just think about it. We have a currency actually is

158
00:08:41,960 --> 00:08:44,400
Speaker 1: more than one. But just to take bitcoin that has

159
00:08:44,400 --> 00:08:46,760
Speaker 1: no government behind it, and a lot of governments are

160
00:08:46,800 --> 00:08:49,560
Speaker 1: quite suspicious of it hasn't been perfect. No, you and

161
00:08:49,600 --> 00:08:52,640
Speaker 1: I can't walk into a store in general and purchase

162
00:08:52,720 --> 00:08:55,480
Speaker 1: coffee with bitcoin, and a lot of criminals use that.

163
00:08:55,520 --> 00:08:58,600
Speaker 1: I understand that. But it's amazing, George, that this thing

164
00:08:58,720 --> 00:09:01,640
Speaker 1: exists at all and persists after ten years and a

165
00:09:01,679 --> 00:09:04,360
Speaker 1: lot of attempts to bring it down. That is pretty

166
00:09:04,440 --> 00:09:08,280
Speaker 1: radical in it promises something that we're severely lacking with

167
00:09:08,320 --> 00:09:11,320
Speaker 1: the Internet and social media. It is amazing. Listen to

168
00:09:11,360 --> 00:09:14,319
Speaker 1: more Coast to Coast a m. Every weeknight at one

169
00:09:14,400 --> 00:09:17,120
Speaker 1: a m. Eastern and go to Coast to Coast am

170
00:09:17,200 --> 00:09:18,200
Speaker 1: dot com for more