WEBVTT
1
00:00:16.550 --> 00:00:21.969
I'm John Batchelor. Artificial intelligence and the Washington leadership. A
2
00:00:22.010 --> 00:00:25.789
headline at the Washington Post tells a story that is
3
00:00:25.989 --> 00:00:29.829
growing in intensity in terms of money, in terms of
4
00:00:29.850 --> 00:00:33.259
the future, in terms of national security. At the White House,
5
00:00:33.700 --> 00:00:37.259
President Trump will pitch an AI, quote, golden age to
6
00:00:37.320 --> 00:00:42.570
a wary nation. Subhead. after rogue AI agents probed agency websites,
7
00:00:43.109 --> 00:00:47.530
Washington Agency. Washington's concerns are at a peak, but a
8
00:00:47.570 --> 00:00:50.590
White House event today will focus on the technology's promise.
9
00:00:51.469 --> 00:00:55.729
This whole story turns very quickly to a discussion of,
10
00:00:55.909 --> 00:01:03.030
are we safe? The large companies, Anthropic and ChatGPI, and
11
00:01:03.079 --> 00:01:06.049
certainly Grok as well, will bring in that, the Mr.
12
00:01:06.079 --> 00:01:09.709
Musk's presentation, there are others, Google and Microsoft, they're all
13
00:01:09.730 --> 00:01:14.150
weighing in for artificial intelligence being not only the future
14
00:01:14.189 --> 00:01:16.450
for the United States, but the future for their profits.
15
00:01:16.870 --> 00:01:20.590
They're spending lavishly with borrowed money to build tech centers,
16
00:01:20.650 --> 00:01:25.409
data centers across the country. They're making large promises, transforming,
17
00:01:25.489 --> 00:01:28.670
I believe, the United States at the scale of the
18
00:01:28.730 --> 00:01:32.469
Industrial Revolution. That was a remark made in these last hours.
19
00:01:33.170 --> 00:01:35.670
At the same time, there's a parallel concern, and I'll
20
00:01:35.709 --> 00:01:41.200
say it's led by the very soft-spoken but serious philosophical
21
00:01:41.359 --> 00:01:45.260
pope himself, saying there's no reason to go so fast.
22
00:01:45.640 --> 00:01:48.719
There are grave concerns here about security for people and
23
00:01:48.819 --> 00:01:54.159
individual respect, and we need guidelines. We need rules. We
24
00:01:54.200 --> 00:01:55.099
need regulations.
25
00:01:55.159 --> 00:01:55.579
Why not?
26
00:01:55.659 --> 00:01:57.939
Why not talk about it? Well, apparently the White House
27
00:01:58.500 --> 00:02:01.140
is talking about it, but Mr. Trump's metaphor of a
28
00:02:01.219 --> 00:02:05.459
golden age suggests that the leader has made up his
29
00:02:05.519 --> 00:02:08.439
mind at once. Jim, I want to begin with you
30
00:02:08.520 --> 00:02:12.840
because Thaddeus and I were discussing your reporting out of
31
00:02:12.900 --> 00:02:15.509
Lancaster County. You live in a county, a prosperous county
32
00:02:15.569 --> 00:02:21.669
of mixed income base, but it's satisfyingly split on the
33
00:02:21.870 --> 00:02:25.289
issue of data centers. Part of the county says no.
34
00:02:25.969 --> 00:02:28.340
A major part of the county says yes. And is
35
00:02:28.379 --> 00:02:32.009
that changing as the argument reaches the front page. Good
36
00:02:32.030 --> 00:02:32.729
day to you, Jim.
37
00:02:33.069 --> 00:02:37.599
Good day. It is changing because more and more municipalities
38
00:02:37.639 --> 00:02:44.139
within the county are adopting regulations that they don't ban AI,
39
00:02:44.199 --> 00:02:51.669
but they circumvent the development. They demand certain construction constraints
40
00:02:51.789 --> 00:02:55.289
on the developers before a data center can go forward.
41
00:02:55.750 --> 00:03:01.270
So I think that's sensible. My position is By all means,
42
00:03:01.330 --> 00:03:05.500
accept the data center, but negotiate. Don't let somebody shove
43
00:03:05.840 --> 00:03:11.219
a facility down your throat. These are deep-pocketed people. And
44
00:03:11.500 --> 00:03:13.550
you can make a deal. If you're worried about noise,
45
00:03:13.590 --> 00:03:19.520
you can have them soundproof all the surrounding housing, etc., So, again,
46
00:03:19.560 --> 00:03:25.080
there are increased restrictions, but there's not blanket rejection, as
47
00:03:25.099 --> 00:03:27.199
you see in some parts of the county and in
48
00:03:27.259 --> 00:03:29.189
some parts of the country. I think in the past
49
00:03:29.210 --> 00:03:34.370
we've talked about Owensboro, Kentucky, which has passed ordinances banning
50
00:03:34.389 --> 00:03:37.810
data centers, and no one ever had proposed a data
51
00:03:37.849 --> 00:03:42.419
center for Owensboro County. Kentucky, there was a rumor that
52
00:03:42.460 --> 00:03:45.340
there was a data center coming, completely false, and they
53
00:03:45.379 --> 00:03:47.949
went ahead and passed a ban. So, I mean, I
54
00:03:47.969 --> 00:03:51.270
think that type of behavior is evident in many parts
55
00:03:51.349 --> 00:03:54.509
of the country and in some parts of Lancaster County.
56
00:03:55.650 --> 00:04:01.020
That is, do you see similar Solomonic decisions about splitting
57
00:04:01.060 --> 00:04:03.080
the baby and slow down in Michigan?
58
00:04:03.099 --> 00:04:06.520
If you talk about data centers, Sean, what we're seeing
59
00:04:06.599 --> 00:04:11.439
is our gubernatorial and our senatorial Democratic nominees are running
60
00:04:11.500 --> 00:04:13.900
against data centers. The concern here, Jim, and I'd be
61
00:04:13.939 --> 00:04:16.990
curious as to what they're saying in Pennsylvania, is that
62
00:04:17.050 --> 00:04:19.350
it's going to be a strain on the utilities, which
63
00:04:19.389 --> 00:04:21.970
will then be passed on to the consumers. And that's
64
00:04:22.029 --> 00:04:24.189
one of the biggest concerns that people have with the
65
00:04:24.230 --> 00:04:27.009
data centers, as opposed to in the past where other
66
00:04:27.889 --> 00:04:30.629
large corporations have put plants in and you've never heard
67
00:04:30.680 --> 00:04:33.300
those types of arguments before. Is that what you're hearing
68
00:04:33.339 --> 00:04:34.040
out of Pennsylvania?
69
00:04:34.060 --> 00:04:38.970
Yeah. I'm actually hearing and I'm looking at ads that
70
00:04:39.009 --> 00:04:43.910
are being run by the Democrats, who the governor previously
71
00:04:44.290 --> 00:04:48.670
embraced data centers until it was unpopular. And now suddenly
72
00:04:48.709 --> 00:04:53.129
he's in favor of restrictions. But now the Democratic ads
73
00:04:53.689 --> 00:04:59.560
paint supporters of data centers as destroyers of the bucolic
74
00:04:59.629 --> 00:05:07.009
Pennsylvania countryside, the destroyers of our environment. So they're wrapping
75
00:05:07.310 --> 00:05:11.509
all those concerns about noise and water and power under
76
00:05:11.550 --> 00:05:13.620
the greeny umbrella.
77
00:05:15.660 --> 00:05:18.939
I think that safety is a major story here, Jim.
78
00:05:19.220 --> 00:05:22.839
I read into the Washington Post coverage of this. And
79
00:05:22.860 --> 00:05:28.670
the subhead here, OpenAI cancels new AI launch citing safety issues.
80
00:05:28.850 --> 00:05:31.230
I'd seen roughly around the edges out of London. This
81
00:05:31.269 --> 00:05:34.300
fills in the details. The move comes days after the
82
00:05:34.339 --> 00:05:38.350
ChatGPT maker said its AI agents probed U.S. government websites.
83
00:05:39.310 --> 00:05:43.550
Here's the detail. The cancellation comes just after OpenAI said
84
00:05:43.589 --> 00:05:47.620
it would pause development of new highly capable AI models
85
00:05:47.680 --> 00:05:50.459
out of concern that the company doesn't have appropriate safeguards
86
00:05:50.500 --> 00:05:54.740
to keep the technology from behaving in unintended ways. AI
87
00:05:54.819 --> 00:05:58.060
companies have made their AI agents, which can take actions
88
00:05:58.139 --> 00:06:01.139
on a computer. more useful to coders and other workers,
89
00:06:01.220 --> 00:06:04.319
in part by making them more persistent in finding solutions
90
00:06:04.899 --> 00:06:09.800
to tasks. It wants to obey the human, but that
91
00:06:09.839 --> 00:06:12.850
has enabled the technology to take on more complex projects,
92
00:06:12.990 --> 00:06:18.129
but can also lead to AI agents that are overzealous
93
00:06:18.350 --> 00:06:22.870
and even cheat to achieve an end goal. Now, some
94
00:06:22.889 --> 00:06:26.870
of this sounds magical. Some of it sounds very organized.
95
00:06:27.860 --> 00:06:31.939
We have regulations on banking. That was necessary, I believe,
96
00:06:31.980 --> 00:06:35.480
in the 19th century because so many banks would disappear overnight.
97
00:06:36.279 --> 00:06:40.139
Why not regulations at that scale on AI? Do you
98
00:06:40.199 --> 00:06:43.060
find an objection to that, Jim? You're a veteran of Barron's.
99
00:06:43.100 --> 00:06:47.240
Would Barron's find an objection to using the bank analogy
100
00:06:47.360 --> 00:06:48.480
to regulate AI?
101
00:06:49.379 --> 00:06:52.170
I can't speak for Barron's anymore because it's been a
102
00:06:52.230 --> 00:06:54.930
long time since I worked there. So I think the
103
00:06:54.970 --> 00:07:00.060
newsroom culture is a little bit different. But in terms
104
00:07:00.120 --> 00:07:02.899
of regulation, I think the debate would be similar to
105
00:07:02.920 --> 00:07:07.959
what we see in financial regulation. Should it be heavy handed, detailed,
106
00:07:08.180 --> 00:07:12.420
lawyered regulation that you see at the Securities and Exchange Commission?
107
00:07:12.740 --> 00:07:17.399
Or should it be a lighter goal oriented regulation that
108
00:07:17.699 --> 00:07:20.720
you see at the Commodities Future Trading Commission? In other words,
109
00:07:21.319 --> 00:07:25.290
should the regulator set a goal and let the industry
110
00:07:25.399 --> 00:07:29.790
figure out how to get there? And the regulators would
111
00:07:29.870 --> 00:07:32.350
look over the shoulder to make sure that they are
112
00:07:32.410 --> 00:07:36.069
making the actual effort as opposed to the Securities and
113
00:07:36.089 --> 00:07:39.829
Exchange Commission, which I pointed out in my book, Crapshoot Investing,
114
00:07:39.850 --> 00:07:44.339
which was about high-frequency trading. I wrote it and it
115
00:07:44.379 --> 00:07:47.160
came out in 2011. That was a form of AI.
116
00:07:48.220 --> 00:07:52.699
The regulators tried to regulate computer trading and they were
117
00:07:52.759 --> 00:07:58.250
so short-sighted and the technology advanced so rapidly that post
118
00:07:58.290 --> 00:08:06.209
their regulation, that the regulators actually created an environment where
119
00:08:07.189 --> 00:08:11.319
stock trading became like casino trading. So my point being
120
00:08:11.339 --> 00:08:15.519
that the regulators are not smart enough to stay on
121
00:08:15.600 --> 00:08:20.139
top of rapidly developing AI. So it's best to use
122
00:08:20.199 --> 00:08:25.490
this commodities future trading commission approach, goal-oriented regulation.
123
00:08:25.899 --> 00:08:28.699
I want to contribute an anecdote to this conversation, gentlemen.
124
00:08:30.060 --> 00:08:34.100
I have entertained myself with one of those trading sites
125
00:08:34.139 --> 00:08:36.940
that are built for young people where there's no cost
126
00:08:36.960 --> 00:08:39.379
to the trading and you buy and sell very quickly.
127
00:08:40.000 --> 00:08:42.120
In my case, I don't have that much time, but
128
00:08:42.539 --> 00:08:44.960
I build up a portfolio, and then one day I
129
00:08:45.019 --> 00:08:49.139
realized the portfolio was reporting an overall number that was
130
00:08:50.519 --> 00:08:52.860
inconsistent with the number I have at my real bank,
131
00:08:53.259 --> 00:08:57.210
a big bank. And every day they would summarize the
132
00:08:57.330 --> 00:08:59.830
assets of the family. And it would come out to
133
00:08:59.870 --> 00:09:02.169
be the wrong number at the bottom. And I puzzled
134
00:09:02.230 --> 00:09:08.100
about this and inquired my banker, my bank manager. I said,
135
00:09:08.129 --> 00:09:11.799
can you explain why it is your number is– I
136
00:09:11.820 --> 00:09:15.460
mean it was considerably off. It was, you know, let's
137
00:09:15.480 --> 00:09:19.240
put it as a five-figure error every time. Okay. So
138
00:09:19.299 --> 00:09:22.740
I said– I don't want to say this is okay.
139
00:09:22.759 --> 00:09:23.529
Can we solve it?
140
00:09:23.960 --> 00:09:24.460
We couldn't.
141
00:09:24.500 --> 00:09:28.629
We kept inquiring, is the feed open? From one banking site,
142
00:09:28.769 --> 00:09:31.529
it was the one called Robinhood, banking site to the
143
00:09:31.570 --> 00:09:34.370
big bank. Yeah, the feed's open. Yeah, it's every day.
144
00:09:35.029 --> 00:09:37.769
So we eventually had to come down to, all right,
145
00:09:37.809 --> 00:09:40.289
let's go to the individual holdings and whether they have
146
00:09:40.309 --> 00:09:42.710
the price right. Turned out the end of the story
147
00:09:42.809 --> 00:09:47.830
is Robinhood was not updating the portfolio. It was just
148
00:09:47.889 --> 00:09:52.029
allowing it to sort of drift in history. And I
149
00:09:52.090 --> 00:09:54.909
thought about it for overnight and realized I don't want
150
00:09:54.950 --> 00:09:57.409
to be with them anymore. A mistake like that is
151
00:09:57.470 --> 00:10:01.789
not small, not at all. And what power do I
152
00:10:01.850 --> 00:10:03.830
have if I want to call up Robin Hood and, hey,
153
00:10:04.129 --> 00:10:05.690
you're five figures off. What are you going to do
154
00:10:05.750 --> 00:10:07.190
about it? And they hung up on me. I wouldn't
155
00:10:07.210 --> 00:10:09.389
have any power at all, except for spend a lot
156
00:10:09.409 --> 00:10:12.289
of money on a lawyer and years waiting. So I
157
00:10:12.330 --> 00:10:15.029
immediately decided I'm out of here and I'm going to
158
00:10:15.210 --> 00:10:18.500
move my account over to the big bank where I'm
159
00:10:18.559 --> 00:10:22.460
more confident in security. Now, I tell that story because
160
00:10:22.519 --> 00:10:26.200
it seems to me that's a version of AI. That
161
00:10:26.279 --> 00:10:27.940
is how I appeal to you. If you see an
162
00:10:28.059 --> 00:10:31.370
error in banking, you want to get out of there.
163
00:10:31.429 --> 00:10:33.529
Isn't that a fair opinion of America?
164
00:10:35.090 --> 00:10:38.289
I think what you're looking at underneath all the concerns
165
00:10:38.340 --> 00:10:42.799
about the data centers is the continuing anxiety of the
166
00:10:42.820 --> 00:10:46.720
American people and people in general. regarding AI and what
167
00:10:46.740 --> 00:10:48.720
its impact upon human beings is going to be. I
168
00:10:48.740 --> 00:10:50.899
think that's all wrapped up into the data centers and
169
00:10:50.940 --> 00:10:54.399
everything else that the Pope was trying to address. And
170
00:10:54.419 --> 00:10:57.159
I think that what's going to happen politically is when
171
00:10:57.179 --> 00:11:00.210
the president comes out in favor of a golden age
172
00:11:00.259 --> 00:11:02.309
of AI, I don't think that's going to do anything
173
00:11:02.330 --> 00:11:05.590
except maybe alleviate the concerns of a handful of partisans
174
00:11:05.610 --> 00:11:07.950
that will be viewed through the partisan lens. I don't
175
00:11:07.970 --> 00:11:09.549
think it's going to do much, but it will leave
176
00:11:09.610 --> 00:11:12.190
Republicans holding the bag on the data center issues as
177
00:11:12.230 --> 00:11:14.779
we head into the midterms. That's going to be a
178
00:11:14.840 --> 00:11:19.259
considerable problem for them. Generally, I think that we do
179
00:11:19.340 --> 00:11:22.809
not know what's going on. I think the AI overlords
180
00:11:23.330 --> 00:11:25.730
want to see regulation on their terms. They don't want
181
00:11:25.769 --> 00:11:28.769
to see a knee-jerk response of heavy-handed regulation, so they're
182
00:11:28.809 --> 00:11:31.409
trying to get it on the ground floor. Basically, with
183
00:11:31.429 --> 00:11:34.250
an argument of stop me before I sin again, I
184
00:11:34.690 --> 00:11:37.149
think Jim is absolutely right. They know more about AI
185
00:11:37.200 --> 00:11:39.379
than anybody else. They've been running around telling people that
186
00:11:39.399 --> 00:11:41.460
it could be the end of the world. So I
187
00:11:41.500 --> 00:11:43.980
think if they really thought that way, they would wind
188
00:11:44.080 --> 00:11:46.980
up slowing down some of their research or stopping it
189
00:11:47.000 --> 00:11:49.519
altogether until they figure out what the hell they're doing.
190
00:11:50.000 --> 00:11:52.059
That's not going to happen because there's too much money
191
00:11:52.139 --> 00:11:54.490
at stake. So we're going to continue to see, at
192
00:11:54.519 --> 00:11:57.590
least in terms of the midterms, the Republicans being left
193
00:11:57.610 --> 00:11:59.950
holding the bag on AI, and that's going to complicate
194
00:11:59.990 --> 00:12:01.669
matters for them to try to hold the Senate.
195
00:12:02.649 --> 00:12:04.909
Let's go to the Pope when we come back. I'm
196
00:12:04.970 --> 00:12:08.250
speaking with Jim McTague, former Washington editor of Barron's Magazine,
197
00:12:08.309 --> 00:12:11.860
now a novelist and reporter in Lancaster County. where there
198
00:12:12.019 --> 00:12:16.830
are two data centers going up and one refused so far.
199
00:12:17.370 --> 00:12:19.710
And Thaddeus McCutter in the great state of Michigan, where
200
00:12:19.769 --> 00:12:24.049
the Senate race has as one of its battle points
201
00:12:24.210 --> 00:12:28.059
the data centers and AI, artificial intelligence. So the future
202
00:12:28.120 --> 00:12:31.600
is here. It's just not evenly distributed. I'm John Batchelor.
203
00:12:50.240 --> 00:12:52.740
I'm John Batchelor with my good colleagues, Jim McTague and
204
00:12:52.779 --> 00:12:56.870
Thaddeus McCotter, and we're discussing, oddly, it all comes together,
205
00:12:56.970 --> 00:13:01.649
data centers, AI, the future, and the election in midterms.
206
00:13:02.210 --> 00:13:05.309
And I go to one more source of inspiration for
207
00:13:05.330 --> 00:13:11.399
me on this story. iRobot, Isaac Asimov, 56th edition, The
208
00:13:11.440 --> 00:13:14.950
Three Laws of Robotics. A robot may not injure a
209
00:13:14.970 --> 00:13:18.190
human being or, through inaction, allow a human being to
210
00:13:18.269 --> 00:13:22.049
come to harm. Second rule. A robot must obey the
211
00:13:22.129 --> 00:13:25.159
orders given it by human beings except where such orders
212
00:13:25.200 --> 00:13:29.220
would conflict with the first law. Third law. A robot
213
00:13:29.259 --> 00:13:32.240
must protect its own existence as long as such protection
214
00:13:32.279 --> 00:13:35.720
does not conflict with the first or second law. So
215
00:13:35.820 --> 00:13:41.159
I offer, gentlemen, the precedent of regulation of iRobot. It
216
00:13:41.220 --> 00:13:44.889
was embedded inhaled by all of us who read science
217
00:13:44.929 --> 00:13:49.019
fiction in our preteen years. And it almost feels like
218
00:13:49.659 --> 00:13:51.980
the White House should at least come up to the
219
00:13:52.059 --> 00:13:55.879
scale of concern that Isaac Asimov once upon a time
220
00:13:56.720 --> 00:13:59.419
issued for his science fiction. How does that strike you, Jim,
221
00:13:59.500 --> 00:14:02.039
as a good starting point, the three laws of robotics?
222
00:14:02.899 --> 00:14:08.879
I think it's an excellent starting point. Um, It isn't
223
00:14:08.889 --> 00:14:11.320
going to happen. We're going to have the three laws
224
00:14:11.399 --> 00:14:19.159
of Trump robotics, I think. And I think it's actually
225
00:14:20.139 --> 00:14:22.700
Trump is the reason that the Pope spoke out, not Jason.
226
00:14:24.600 --> 00:14:28.559
Let's go to the Pope's remarks. The Pope said this
227
00:14:28.659 --> 00:14:33.710
is a National Catholic Register reporting report. Pope Leo waded
228
00:14:33.769 --> 00:14:37.690
into the red-hot debate over regulation of artificial intelligence, stating
229
00:14:37.730 --> 00:14:42.080
that concerns expressed by tech industry leaders over the catastrophic
230
00:14:42.129 --> 00:14:47.039
effects risk to humanity posed by technology are not, not
231
00:14:47.179 --> 00:14:50.679
fake news, said the Pope. And they should be taken seriously.
232
00:14:50.840 --> 00:14:52.419
He went on to say, this is a problem that
233
00:14:52.480 --> 00:14:55.320
we need to sit down and talk about. And that's
234
00:14:55.340 --> 00:14:56.960
what you're referring to. Is that correct?
235
00:14:57.080 --> 00:15:00.360
Yes. When he used the term fake news to introduce
236
00:15:00.419 --> 00:15:05.559
his remarks, I mean, that's... That's a Trump brand label.
237
00:15:06.500 --> 00:15:10.519
And the Pope is in Europe where Trump is a
238
00:15:11.070 --> 00:15:17.269
source of dread and distaste, and he's become fodder for humor.
239
00:15:18.210 --> 00:15:23.789
There's a French comedy called Call My Agent, which is hilarious.
240
00:15:23.850 --> 00:15:27.470
It's on HBO now. But in one scene, they're trying
241
00:15:27.490 --> 00:15:30.870
to recruit an actor who is in an asylum recovering
242
00:15:30.889 --> 00:15:34.750
from from a nervous breakdown caused by anxiety. And when
243
00:15:34.769 --> 00:15:39.029
the actor is listing the sources of anxiety, he matter
244
00:15:39.250 --> 00:15:42.899
of factly tops the list with Trump, which is a
245
00:15:43.100 --> 00:15:47.539
hilarious line, but I think it summarizes, uh, the way
246
00:15:47.600 --> 00:15:51.259
our president is seen in Europe, a source of distaste
247
00:15:51.539 --> 00:15:54.279
and dread. And so when he gathers with the head
248
00:15:54.320 --> 00:16:04.100
of all these AI agents to promote, uh, non-regulation, globally,
249
00:16:04.600 --> 00:16:07.820
it creates unease and disbelief.
250
00:16:08.679 --> 00:16:11.480
That is, how do you see the Pope's remarks? A
251
00:16:11.659 --> 00:16:15.519
necessary guideline? Do you think they'll be effective at holding
252
00:16:15.580 --> 00:16:20.379
the White House back from the golden age of AI?
253
00:16:21.740 --> 00:16:25.120
I think that the Pope has long been talking about AI.
254
00:16:25.779 --> 00:16:28.210
He's been talking about it for quite some time in
255
00:16:28.230 --> 00:16:32.950
the sense of theological the philosophical and the moral dimensions
256
00:16:33.029 --> 00:16:35.830
of it and its impact upon humans. I think he
257
00:16:35.870 --> 00:16:38.049
may have used the term fake news to make sure
258
00:16:38.070 --> 00:16:42.190
that he pierces through the din out there from the
259
00:16:42.250 --> 00:16:47.820
social networks. But in the final analysis, that's the perception
260
00:16:47.860 --> 00:16:50.240
he's coming from. That's the perspective he's coming from. I
261
00:16:50.259 --> 00:16:53.220
think it's the proper perspective to look at this. Before
262
00:16:53.279 --> 00:16:55.720
you get into any issues about regulation at all, there
263
00:16:55.740 --> 00:16:58.039
has to be a a firm understanding of what it
264
00:16:58.059 --> 00:16:59.960
can or cannot do in the sense of how it
265
00:17:00.000 --> 00:17:03.559
impacts human beings, how it affects our innate dignity, our
266
00:17:03.580 --> 00:17:07.660
innate capacity, and our innate agency as human beings. That's
267
00:17:07.680 --> 00:17:10.670
the perspective he's coming from, obviously, of course, as children
268
00:17:10.710 --> 00:17:14.309
of God, how it affects our relationship with Christ. So
269
00:17:14.349 --> 00:17:17.210
that's where he's coming from. It's a far deeper, I think,
270
00:17:18.990 --> 00:17:23.789
far deeper understanding and questioning of AI than you're seeing
271
00:17:23.829 --> 00:17:25.309
out of the political sector.
272
00:17:26.349 --> 00:17:30.140
That is McCutter, American Greatness, Jim McTague, former Washington editor
273
00:17:30.160 --> 00:17:34.049
of Barron's, now a novelist in Lancaster County, Pennsylvania, the
274
00:17:34.069 --> 00:17:36.410
Keystone State. I'm John Batchelor.
1
00:00:16.550 --> 00:00:21.969
I'm John Batchelor. Artificial intelligence and the Washington leadership. A
2
00:00:22.010 --> 00:00:25.789
headline at the Washington Post tells a story that is
3
00:00:25.989 --> 00:00:29.829
growing in intensity in terms of money, in terms of
4
00:00:29.850 --> 00:00:33.259
the future, in terms of national security. At the White House,
5
00:00:33.700 --> 00:00:37.259
President Trump will pitch an AI, quote, golden age to
6
00:00:37.320 --> 00:00:42.570
a wary nation. Subhead. after rogue AI agents probed agency websites,
7
00:00:43.109 --> 00:00:47.530
Washington Agency. Washington's concerns are at a peak, but a
8
00:00:47.570 --> 00:00:50.590
White House event today will focus on the technology's promise.
9
00:00:51.469 --> 00:00:55.729
This whole story turns very quickly to a discussion of,
10
00:00:55.909 --> 00:01:03.030
are we safe? The large companies, Anthropic and ChatGPI, and
11
00:01:03.079 --> 00:01:06.049
certainly Grok as well, will bring in that, the Mr.
12
00:01:06.079 --> 00:01:09.709
Musk's presentation, there are others, Google and Microsoft, they're all
13
00:01:09.730 --> 00:01:14.150
weighing in for artificial intelligence being not only the future
14
00:01:14.189 --> 00:01:16.450
for the United States, but the future for their profits.
15
00:01:16.870 --> 00:01:20.590
They're spending lavishly with borrowed money to build tech centers,
16
00:01:20.650 --> 00:01:25.409
data centers across the country. They're making large promises, transforming,
17
00:01:25.489 --> 00:01:28.670
I believe, the United States at the scale of the
18
00:01:28.730 --> 00:01:32.469
Industrial Revolution. That was a remark made in these last hours.
19
00:01:33.170 --> 00:01:35.670
At the same time, there's a parallel concern, and I'll
20
00:01:35.709 --> 00:01:41.200
say it's led by the very soft-spoken but serious philosophical
21
00:01:41.359 --> 00:01:45.260
pope himself, saying there's no reason to go so fast.
22
00:01:45.640 --> 00:01:48.719
There are grave concerns here about security for people and
23
00:01:48.819 --> 00:01:54.159
individual respect, and we need guidelines. We need rules. We
24
00:01:54.200 --> 00:01:55.099
need regulations.
25
00:01:55.159 --> 00:01:55.579
Why not?
26
00:01:55.659 --> 00:01:57.939
Why not talk about it? Well, apparently the White House
27
00:01:58.500 --> 00:02:01.140
is talking about it, but Mr. Trump's metaphor of a
28
00:02:01.219 --> 00:02:05.459
golden age suggests that the leader has made up his
29
00:02:05.519 --> 00:02:08.439
mind at once. Jim, I want to begin with you
30
00:02:08.520 --> 00:02:12.840
because Thaddeus and I were discussing your reporting out of
31
00:02:12.900 --> 00:02:15.509
Lancaster County. You live in a county, a prosperous county
32
00:02:15.569 --> 00:02:21.669
of mixed income base, but it's satisfyingly split on the
33
00:02:21.870 --> 00:02:25.289
issue of data centers. Part of the county says no.
34
00:02:25.969 --> 00:02:28.340
A major part of the county says yes. And is
35
00:02:28.379 --> 00:02:32.009
that changing as the argument reaches the front page. Good
36
00:02:32.030 --> 00:02:32.729
day to you, Jim.
37
00:02:33.069 --> 00:02:37.599
Good day. It is changing because more and more municipalities
38
00:02:37.639 --> 00:02:44.139
within the county are adopting regulations that they don't ban AI,
39
00:02:44.199 --> 00:02:51.669
but they circumvent the development. They demand certain construction constraints
40
00:02:51.789 --> 00:02:55.289
on the developers before a data center can go forward.
41
00:02:55.750 --> 00:03:01.270
So I think that's sensible. My position is By all means,
42
00:03:01.330 --> 00:03:05.500
accept the data center, but negotiate. Don't let somebody shove
43
00:03:05.840 --> 00:03:11.219
a facility down your throat. These are deep-pocketed people. And
44
00:03:11.500 --> 00:03:13.550
you can make a deal. If you're worried about noise,
45
00:03:13.590 --> 00:03:19.520
you can have them soundproof all the surrounding housing, etc., So, again,
46
00:03:19.560 --> 00:03:25.080
there are increased restrictions, but there's not blanket rejection, as
47
00:03:25.099 --> 00:03:27.199
you see in some parts of the county and in
48
00:03:27.259 --> 00:03:29.189
some parts of the country. I think in the past
49
00:03:29.210 --> 00:03:34.370
we've talked about Owensboro, Kentucky, which has passed ordinances banning
50
00:03:34.389 --> 00:03:37.810
data centers, and no one ever had proposed a data
51
00:03:37.849 --> 00:03:42.419
center for Owensboro County. Kentucky, there was a rumor that
52
00:03:42.460 --> 00:03:45.340
there was a data center coming, completely false, and they
53
00:03:45.379 --> 00:03:47.949
went ahead and passed a ban. So, I mean, I
54
00:03:47.969 --> 00:03:51.270
think that type of behavior is evident in many parts
55
00:03:51.349 --> 00:03:54.509
of the country and in some parts of Lancaster County.
56
00:03:55.650 --> 00:04:01.020
That is, do you see similar Solomonic decisions about splitting
57
00:04:01.060 --> 00:04:03.080
the baby and slow down in Michigan?
58
00:04:03.099 --> 00:04:06.520
If you talk about data centers, Sean, what we're seeing
59
00:04:06.599 --> 00:04:11.439
is our gubernatorial and our senatorial Democratic nominees are running
60
00:04:11.500 --> 00:04:13.900
against data centers. The concern here, Jim, and I'd be
61
00:04:13.939 --> 00:04:16.990
curious as to what they're saying in Pennsylvania, is that
62
00:04:17.050 --> 00:04:19.350
it's going to be a strain on the utilities, which
63
00:04:19.389 --> 00:04:21.970
will then be passed on to the consumers. And that's
64
00:04:22.029 --> 00:04:24.189
one of the biggest concerns that people have with the
65
00:04:24.230 --> 00:04:27.009
data centers, as opposed to in the past where other
66
00:04:27.889 --> 00:04:30.629
large corporations have put plants in and you've never heard
67
00:04:30.680 --> 00:04:33.300
those types of arguments before. Is that what you're hearing
68
00:04:33.339 --> 00:04:34.040
out of Pennsylvania?
69
00:04:34.060 --> 00:04:38.970
Yeah. I'm actually hearing and I'm looking at ads that
70
00:04:39.009 --> 00:04:43.910
are being run by the Democrats, who the governor previously
71
00:04:44.290 --> 00:04:48.670
embraced data centers until it was unpopular. And now suddenly
72
00:04:48.709 --> 00:04:53.129
he's in favor of restrictions. But now the Democratic ads
73
00:04:53.689 --> 00:04:59.560
paint supporters of data centers as destroyers of the bucolic
74
00:04:59.629 --> 00:05:07.009
Pennsylvania countryside, the destroyers of our environment. So they're wrapping
75
00:05:07.310 --> 00:05:11.509
all those concerns about noise and water and power under
76
00:05:11.550 --> 00:05:13.620
the greeny umbrella.
77
00:05:15.660 --> 00:05:18.939
I think that safety is a major story here, Jim.
78
00:05:19.220 --> 00:05:22.839
I read into the Washington Post coverage of this. And
79
00:05:22.860 --> 00:05:28.670
the subhead here, OpenAI cancels new AI launch citing safety issues.
80
00:05:28.850 --> 00:05:31.230
I'd seen roughly around the edges out of London. This
81
00:05:31.269 --> 00:05:34.300
fills in the details. The move comes days after the
82
00:05:34.339 --> 00:05:38.350
ChatGPT maker said its AI agents probed U.S. government websites.
83
00:05:39.310 --> 00:05:43.550
Here's the detail. The cancellation comes just after OpenAI said
84
00:05:43.589 --> 00:05:47.620
it would pause development of new highly capable AI models
85
00:05:47.680 --> 00:05:50.459
out of concern that the company doesn't have appropriate safeguards
86
00:05:50.500 --> 00:05:54.740
to keep the technology from behaving in unintended ways. AI
87
00:05:54.819 --> 00:05:58.060
companies have made their AI agents, which can take actions
88
00:05:58.139 --> 00:06:01.139
on a computer. more useful to coders and other workers,
89
00:06:01.220 --> 00:06:04.319
in part by making them more persistent in finding solutions
90
00:06:04.899 --> 00:06:09.800
to tasks. It wants to obey the human, but that
91
00:06:09.839 --> 00:06:12.850
has enabled the technology to take on more complex projects,
92
00:06:12.990 --> 00:06:18.129
but can also lead to AI agents that are overzealous
93
00:06:18.350 --> 00:06:22.870
and even cheat to achieve an end goal. Now, some
94
00:06:22.889 --> 00:06:26.870
of this sounds magical. Some of it sounds very organized.
95
00:06:27.860 --> 00:06:31.939
We have regulations on banking. That was necessary, I believe,
96
00:06:31.980 --> 00:06:35.480
in the 19th century because so many banks would disappear overnight.
97
00:06:36.279 --> 00:06:40.139
Why not regulations at that scale on AI? Do you
98
00:06:40.199 --> 00:06:43.060
find an objection to that, Jim? You're a veteran of Barron's.
99
00:06:43.100 --> 00:06:47.240
Would Barron's find an objection to using the bank analogy
100
00:06:47.360 --> 00:06:48.480
to regulate AI?
101
00:06:49.379 --> 00:06:52.170
I can't speak for Barron's anymore because it's been a
102
00:06:52.230 --> 00:06:54.930
long time since I worked there. So I think the
103
00:06:54.970 --> 00:07:00.060
newsroom culture is a little bit different. But in terms
104
00:07:00.120 --> 00:07:02.899
of regulation, I think the debate would be similar to
105
00:07:02.920 --> 00:07:07.959
what we see in financial regulation. Should it be heavy handed, detailed,
106
00:07:08.180 --> 00:07:12.420
lawyered regulation that you see at the Securities and Exchange Commission?
107
00:07:12.740 --> 00:07:17.399
Or should it be a lighter goal oriented regulation that
108
00:07:17.699 --> 00:07:20.720
you see at the Commodities Future Trading Commission? In other words,
109
00:07:21.319 --> 00:07:25.290
should the regulator set a goal and let the industry
110
00:07:25.399 --> 00:07:29.790
figure out how to get there? And the regulators would
111
00:07:29.870 --> 00:07:32.350
look over the shoulder to make sure that they are
112
00:07:32.410 --> 00:07:36.069
making the actual effort as opposed to the Securities and
113
00:07:36.089 --> 00:07:39.829
Exchange Commission, which I pointed out in my book, Crapshoot Investing,
114
00:07:39.850 --> 00:07:44.339
which was about high-frequency trading. I wrote it and it
115
00:07:44.379 --> 00:07:47.160
came out in 2011. That was a form of AI.
116
00:07:48.220 --> 00:07:52.699
The regulators tried to regulate computer trading and they were
117
00:07:52.759 --> 00:07:58.250
so short-sighted and the technology advanced so rapidly that post
118
00:07:58.290 --> 00:08:06.209
their regulation, that the regulators actually created an environment where
119
00:08:07.189 --> 00:08:11.319
stock trading became like casino trading. So my point being
120
00:08:11.339 --> 00:08:15.519
that the regulators are not smart enough to stay on
121
00:08:15.600 --> 00:08:20.139
top of rapidly developing AI. So it's best to use
122
00:08:20.199 --> 00:08:25.490
this commodities future trading commission approach, goal-oriented regulation.
123
00:08:25.899 --> 00:08:28.699
I want to contribute an anecdote to this conversation, gentlemen.
124
00:08:30.060 --> 00:08:34.100
I have entertained myself with one of those trading sites
125
00:08:34.139 --> 00:08:36.940
that are built for young people where there's no cost
126
00:08:36.960 --> 00:08:39.379
to the trading and you buy and sell very quickly.
127
00:08:40.000 --> 00:08:42.120
In my case, I don't have that much time, but
128
00:08:42.539 --> 00:08:44.960
I build up a portfolio, and then one day I
129
00:08:45.019 --> 00:08:49.139
realized the portfolio was reporting an overall number that was
130
00:08:50.519 --> 00:08:52.860
inconsistent with the number I have at my real bank,
131
00:08:53.259 --> 00:08:57.210
a big bank. And every day they would summarize the
132
00:08:57.330 --> 00:08:59.830
assets of the family. And it would come out to
133
00:08:59.870 --> 00:09:02.169
be the wrong number at the bottom. And I puzzled
134
00:09:02.230 --> 00:09:08.100
about this and inquired my banker, my bank manager. I said,
135
00:09:08.129 --> 00:09:11.799
can you explain why it is your number is– I
136
00:09:11.820 --> 00:09:15.460
mean it was considerably off. It was, you know, let's
137
00:09:15.480 --> 00:09:19.240
put it as a five-figure error every time. Okay. So
138
00:09:19.299 --> 00:09:22.740
I said– I don't want to say this is okay.
139
00:09:22.759 --> 00:09:23.529
Can we solve it?
140
00:09:23.960 --> 00:09:24.460
We couldn't.
141
00:09:24.500 --> 00:09:28.629
We kept inquiring, is the feed open? From one banking site,
142
00:09:28.769 --> 00:09:31.529
it was the one called Robinhood, banking site to the
143
00:09:31.570 --> 00:09:34.370
big bank. Yeah, the feed's open. Yeah, it's every day.
144
00:09:35.029 --> 00:09:37.769
So we eventually had to come down to, all right,
145
00:09:37.809 --> 00:09:40.289
let's go to the individual holdings and whether they have
146
00:09:40.309 --> 00:09:42.710
the price right. Turned out the end of the story
147
00:09:42.809 --> 00:09:47.830
is Robinhood was not updating the portfolio. It was just
148
00:09:47.889 --> 00:09:52.029
allowing it to sort of drift in history. And I
149
00:09:52.090 --> 00:09:54.909
thought about it for overnight and realized I don't want
150
00:09:54.950 --> 00:09:57.409
to be with them anymore. A mistake like that is
151
00:09:57.470 --> 00:10:01.789
not small, not at all. And what power do I
152
00:10:01.850 --> 00:10:03.830
have if I want to call up Robin Hood and, hey,
153
00:10:04.129 --> 00:10:05.690
you're five figures off. What are you going to do
154
00:10:05.750 --> 00:10:07.190
about it? And they hung up on me. I wouldn't
155
00:10:07.210 --> 00:10:09.389
have any power at all, except for spend a lot
156
00:10:09.409 --> 00:10:12.289
of money on a lawyer and years waiting. So I
157
00:10:12.330 --> 00:10:15.029
immediately decided I'm out of here and I'm going to
158
00:10:15.210 --> 00:10:18.500
move my account over to the big bank where I'm
159
00:10:18.559 --> 00:10:22.460
more confident in security. Now, I tell that story because
160
00:10:22.519 --> 00:10:26.200
it seems to me that's a version of AI. That
161
00:10:26.279 --> 00:10:27.940
is how I appeal to you. If you see an
162
00:10:28.059 --> 00:10:31.370
error in banking, you want to get out of there.
163
00:10:31.429 --> 00:10:33.529
Isn't that a fair opinion of America?
164
00:10:35.090 --> 00:10:38.289
I think what you're looking at underneath all the concerns
165
00:10:38.340 --> 00:10:42.799
about the data centers is the continuing anxiety of the
166
00:10:42.820 --> 00:10:46.720
American people and people in general. regarding AI and what
167
00:10:46.740 --> 00:10:48.720
its impact upon human beings is going to be. I
168
00:10:48.740 --> 00:10:50.899
think that's all wrapped up into the data centers and
169
00:10:50.940 --> 00:10:54.399
everything else that the Pope was trying to address. And
170
00:10:54.419 --> 00:10:57.159
I think that what's going to happen politically is when
171
00:10:57.179 --> 00:11:00.210
the president comes out in favor of a golden age
172
00:11:00.259 --> 00:11:02.309
of AI, I don't think that's going to do anything
173
00:11:02.330 --> 00:11:05.590
except maybe alleviate the concerns of a handful of partisans
174
00:11:05.610 --> 00:11:07.950
that will be viewed through the partisan lens. I don't
175
00:11:07.970 --> 00:11:09.549
think it's going to do much, but it will leave
176
00:11:09.610 --> 00:11:12.190
Republicans holding the bag on the data center issues as
177
00:11:12.230 --> 00:11:14.779
we head into the midterms. That's going to be a
178
00:11:14.840 --> 00:11:19.259
considerable problem for them. Generally, I think that we do
179
00:11:19.340 --> 00:11:22.809
not know what's going on. I think the AI overlords
180
00:11:23.330 --> 00:11:25.730
want to see regulation on their terms. They don't want
181
00:11:25.769 --> 00:11:28.769
to see a knee-jerk response of heavy-handed regulation, so they're
182
00:11:28.809 --> 00:11:31.409
trying to get it on the ground floor. Basically, with
183
00:11:31.429 --> 00:11:34.250
an argument of stop me before I sin again, I
184
00:11:34.690 --> 00:11:37.149
think Jim is absolutely right. They know more about AI
185
00:11:37.200 --> 00:11:39.379
than anybody else. They've been running around telling people that
186
00:11:39.399 --> 00:11:41.460
it could be the end of the world. So I
187
00:11:41.500 --> 00:11:43.980
think if they really thought that way, they would wind
188
00:11:44.080 --> 00:11:46.980
up slowing down some of their research or stopping it
189
00:11:47.000 --> 00:11:49.519
altogether until they figure out what the hell they're doing.
190
00:11:50.000 --> 00:11:52.059
That's not going to happen because there's too much money
191
00:11:52.139 --> 00:11:54.490
at stake. So we're going to continue to see, at
192
00:11:54.519 --> 00:11:57.590
least in terms of the midterms, the Republicans being left
193
00:11:57.610 --> 00:11:59.950
holding the bag on AI, and that's going to complicate
194
00:11:59.990 --> 00:12:01.669
matters for them to try to hold the Senate.
195
00:12:02.649 --> 00:12:04.909
Let's go to the Pope when we come back. I'm
196
00:12:04.970 --> 00:12:08.250
speaking with Jim McTague, former Washington editor of Barron's Magazine,
197
00:12:08.309 --> 00:12:11.860
now a novelist and reporter in Lancaster County. where there
198
00:12:12.019 --> 00:12:16.830
are two data centers going up and one refused so far.
199
00:12:17.370 --> 00:12:19.710
And Thaddeus McCutter in the great state of Michigan, where
200
00:12:19.769 --> 00:12:24.049
the Senate race has as one of its battle points
201
00:12:24.210 --> 00:12:28.059
the data centers and AI, artificial intelligence. So the future
202
00:12:28.120 --> 00:12:31.600
is here. It's just not evenly distributed. I'm John Batchelor.
203
00:12:50.240 --> 00:12:52.740
I'm John Batchelor with my good colleagues, Jim McTague and
204
00:12:52.779 --> 00:12:56.870
Thaddeus McCotter, and we're discussing, oddly, it all comes together,
205
00:12:56.970 --> 00:13:01.649
data centers, AI, the future, and the election in midterms.
206
00:13:02.210 --> 00:13:05.309
And I go to one more source of inspiration for
207
00:13:05.330 --> 00:13:11.399
me on this story. iRobot, Isaac Asimov, 56th edition, The
208
00:13:11.440 --> 00:13:14.950
Three Laws of Robotics. A robot may not injure a
209
00:13:14.970 --> 00:13:18.190
human being or, through inaction, allow a human being to
210
00:13:18.269 --> 00:13:22.049
come to harm. Second rule. A robot must obey the
211
00:13:22.129 --> 00:13:25.159
orders given it by human beings except where such orders
212
00:13:25.200 --> 00:13:29.220
would conflict with the first law. Third law. A robot
213
00:13:29.259 --> 00:13:32.240
must protect its own existence as long as such protection
214
00:13:32.279 --> 00:13:35.720
does not conflict with the first or second law. So
215
00:13:35.820 --> 00:13:41.159
I offer, gentlemen, the precedent of regulation of iRobot. It
216
00:13:41.220 --> 00:13:44.889
was embedded inhaled by all of us who read science
217
00:13:44.929 --> 00:13:49.019
fiction in our preteen years. And it almost feels like
218
00:13:49.659 --> 00:13:51.980
the White House should at least come up to the
219
00:13:52.059 --> 00:13:55.879
scale of concern that Isaac Asimov once upon a time
220
00:13:56.720 --> 00:13:59.419
issued for his science fiction. How does that strike you, Jim,
221
00:13:59.500 --> 00:14:02.039
as a good starting point, the three laws of robotics?
222
00:14:02.899 --> 00:14:08.879
I think it's an excellent starting point. Um, It isn't
223
00:14:08.889 --> 00:14:11.320
going to happen. We're going to have the three laws
224
00:14:11.399 --> 00:14:19.159
of Trump robotics, I think. And I think it's actually
225
00:14:20.139 --> 00:14:22.700
Trump is the reason that the Pope spoke out, not Jason.
226
00:14:24.600 --> 00:14:28.559
Let's go to the Pope's remarks. The Pope said this
227
00:14:28.659 --> 00:14:33.710
is a National Catholic Register reporting report. Pope Leo waded
228
00:14:33.769 --> 00:14:37.690
into the red-hot debate over regulation of artificial intelligence, stating
229
00:14:37.730 --> 00:14:42.080
that concerns expressed by tech industry leaders over the catastrophic
230
00:14:42.129 --> 00:14:47.039
effects risk to humanity posed by technology are not, not
231
00:14:47.179 --> 00:14:50.679
fake news, said the Pope. And they should be taken seriously.
232
00:14:50.840 --> 00:14:52.419
He went on to say, this is a problem that
233
00:14:52.480 --> 00:14:55.320
we need to sit down and talk about. And that's
234
00:14:55.340 --> 00:14:56.960
what you're referring to. Is that correct?
235
00:14:57.080 --> 00:15:00.360
Yes. When he used the term fake news to introduce
236
00:15:00.419 --> 00:15:05.559
his remarks, I mean, that's... That's a Trump brand label.
237
00:15:06.500 --> 00:15:10.519
And the Pope is in Europe where Trump is a
238
00:15:11.070 --> 00:15:17.269
source of dread and distaste, and he's become fodder for humor.
239
00:15:18.210 --> 00:15:23.789
There's a French comedy called Call My Agent, which is hilarious.
240
00:15:23.850 --> 00:15:27.470
It's on HBO now. But in one scene, they're trying
241
00:15:27.490 --> 00:15:30.870
to recruit an actor who is in an asylum recovering
242
00:15:30.889 --> 00:15:34.750
from from a nervous breakdown caused by anxiety. And when
243
00:15:34.769 --> 00:15:39.029
the actor is listing the sources of anxiety, he matter
244
00:15:39.250 --> 00:15:42.899
of factly tops the list with Trump, which is a
245
00:15:43.100 --> 00:15:47.539
hilarious line, but I think it summarizes, uh, the way
246
00:15:47.600 --> 00:15:51.259
our president is seen in Europe, a source of distaste
247
00:15:51.539 --> 00:15:54.279
and dread. And so when he gathers with the head
248
00:15:54.320 --> 00:16:04.100
of all these AI agents to promote, uh, non-regulation, globally,
249
00:16:04.600 --> 00:16:07.820
it creates unease and disbelief.
250
00:16:08.679 --> 00:16:11.480
That is, how do you see the Pope's remarks? A
251
00:16:11.659 --> 00:16:15.519
necessary guideline? Do you think they'll be effective at holding
252
00:16:15.580 --> 00:16:20.379
the White House back from the golden age of AI?
253
00:16:21.740 --> 00:16:25.120
I think that the Pope has long been talking about AI.
254
00:16:25.779 --> 00:16:28.210
He's been talking about it for quite some time in
255
00:16:28.230 --> 00:16:32.950
the sense of theological the philosophical and the moral dimensions
256
00:16:33.029 --> 00:16:35.830
of it and its impact upon humans. I think he
257
00:16:35.870 --> 00:16:38.049
may have used the term fake news to make sure
258
00:16:38.070 --> 00:16:42.190
that he pierces through the din out there from the
259
00:16:42.250 --> 00:16:47.820
social networks. But in the final analysis, that's the perception
260
00:16:47.860 --> 00:16:50.240
he's coming from. That's the perspective he's coming from. I
261
00:16:50.259 --> 00:16:53.220
think it's the proper perspective to look at this. Before
262
00:16:53.279 --> 00:16:55.720
you get into any issues about regulation at all, there
263
00:16:55.740 --> 00:16:58.039
has to be a a firm understanding of what it
264
00:16:58.059 --> 00:16:59.960
can or cannot do in the sense of how it
265
00:17:00.000 --> 00:17:03.559
impacts human beings, how it affects our innate dignity, our
266
00:17:03.580 --> 00:17:07.660
innate capacity, and our innate agency as human beings. That's
267
00:17:07.680 --> 00:17:10.670
the perspective he's coming from, obviously, of course, as children
268
00:17:10.710 --> 00:17:14.309
of God, how it affects our relationship with Christ. So
269
00:17:14.349 --> 00:17:17.210
that's where he's coming from. It's a far deeper, I think,
270
00:17:18.990 --> 00:17:23.789
far deeper understanding and questioning of AI than you're seeing
271
00:17:23.829 --> 00:17:25.309
out of the political sector.
272
00:17:26.349 --> 00:17:30.140
That is McCutter, American Greatness, Jim McTague, former Washington editor
273
00:17:30.160 --> 00:17:34.049
of Barron's, now a novelist in Lancaster County, Pennsylvania, the
274
00:17:34.069 --> 00:17:36.410
Keystone State. I'm John Batchelor.
