WEBVTT
Kind: captions
Language: en

00:00:00.000 --> 00:00:01.750 align:start position:0%
 
Why<00:00:00.240><c> are</c><00:00:00.320><c> we</c><00:00:00.440><c> trying</c><00:00:00.680><c> to</c><00:00:00.800><c> build</c><00:00:01.160><c> AI</c><00:00:01.400><c> systems</c>

00:00:01.750 --> 00:00:01.760 align:start position:0%
Why are we trying to build AI systems
 

00:00:01.760 --> 00:00:04.070 align:start position:0%
Why are we trying to build AI systems
that<00:00:01.880><c> are</c><00:00:01.960><c> duplicative</c><00:00:02.600><c> of</c><00:00:02.720><c> humans?</c><00:00:03.800><c> They</c>

00:00:04.070 --> 00:00:04.080 align:start position:0%
that are duplicative of humans? They
 

00:00:04.080 --> 00:00:05.870 align:start position:0%
that are duplicative of humans? They
said<00:00:04.360><c> that</c><00:00:04.520><c> we</c><00:00:04.640><c> should</c><00:00:04.800><c> be</c><00:00:04.880><c> building</c><00:00:05.280><c> AGI,</c><00:00:05.720><c> so</c>

00:00:05.870 --> 00:00:05.880 align:start position:0%
said that we should be building AGI, so
 

00:00:05.880 --> 00:00:07.550 align:start position:0%
said that we should be building AGI, so
we<00:00:06.080><c> say</c><00:00:06.280><c> that</c><00:00:06.440><c> we</c><00:00:06.560><c> should</c><00:00:06.680><c> be</c><00:00:06.760><c> building</c><00:00:07.120><c> AGI,</c>

00:00:07.550 --> 00:00:07.560 align:start position:0%
we say that we should be building AGI,
 

00:00:07.560 --> 00:00:09.390 align:start position:0%
we say that we should be building AGI,
but<00:00:07.880><c> I</c><00:00:08.000><c> would</c><00:00:08.120><c> like</c><00:00:08.280><c> to</c><00:00:08.400><c> ask</c><00:00:08.720><c> like</c><00:00:08.920><c> why</c><00:00:09.120><c> are</c><00:00:09.240><c> we</c>

00:00:09.390 --> 00:00:09.400 align:start position:0%
but I would like to ask like why are we
 

00:00:09.400 --> 00:00:12.230 align:start position:0%
but I would like to ask like why are we
doing<00:00:09.680><c> that?</c><00:00:10.400><c> Why</c><00:00:10.920><c> is</c><00:00:11.200><c> it</c><00:00:11.560><c> that</c><00:00:12.000><c> we</c><00:00:12.120><c> are</c>

00:00:12.230 --> 00:00:12.240 align:start position:0%
doing that? Why is it that we are
 

00:00:12.240 --> 00:00:14.670 align:start position:0%
doing that? Why is it that we are
building<00:00:12.600><c> a</c><00:00:12.680><c> technology</c><00:00:13.640><c> that</c><00:00:13.840><c> is</c><00:00:14.080><c> ultimately</c>

00:00:14.670 --> 00:00:14.680 align:start position:0%
building a technology that is ultimately
 

00:00:14.680 --> 00:00:17.150 align:start position:0%
building a technology that is ultimately
designed<00:00:15.240><c> to</c><00:00:15.360><c> replace</c><00:00:15.920><c> and</c><00:00:16.120><c> automate</c><00:00:16.720><c> people</c>

00:00:17.150 --> 00:00:17.160 align:start position:0%
designed to replace and automate people
 

00:00:17.160 --> 00:00:19.750 align:start position:0%
designed to replace and automate people
away?<00:00:18.080><c> We</c><00:00:18.320><c> should</c><00:00:18.520><c> be</c><00:00:18.680><c> building</c><00:00:19.040><c> technology</c>

00:00:19.750 --> 00:00:19.760 align:start position:0%
away? We should be building technology
 

00:00:19.760 --> 00:00:21.710 align:start position:0%
away? We should be building technology
and<00:00:20.160><c> the</c><00:00:20.280><c> purpose</c><00:00:20.720><c> of</c><00:00:20.840><c> technology</c><00:00:21.360><c> throughout</c>

00:00:21.710 --> 00:00:21.720 align:start position:0%
and the purpose of technology throughout
 

00:00:21.720 --> 00:00:24.430 align:start position:0%
and the purpose of technology throughout
history<00:00:22.560><c> has</c><00:00:22.760><c> been</c><00:00:22.920><c> to</c><00:00:23.040><c> improve</c><00:00:24.080><c> human</c>

00:00:24.430 --> 00:00:24.440 align:start position:0%
history has been to improve human
 

00:00:24.440 --> 00:00:28.310 align:start position:0%
history has been to improve human
flourishing,<00:00:25.400><c> not</c><00:00:25.880><c> to</c><00:00:26.280><c> replace</c><00:00:27.040><c> people.</c><00:00:28.080><c> Is</c>

00:00:28.310 --> 00:00:28.320 align:start position:0%
flourishing, not to replace people. Is
 

00:00:28.320 --> 00:00:30.270 align:start position:0%
flourishing, not to replace people. Is
this<00:00:28.520><c> the</c><00:00:28.640><c> right</c><00:00:28.880><c> goal?</c><00:00:29.400><c> What</c><00:00:29.680><c> like</c><00:00:29.960><c> why</c><00:00:30.160><c> are</c>

00:00:30.270 --> 00:00:30.280 align:start position:0%
this the right goal? What like why are
 

00:00:30.280 --> 00:00:33.150 align:start position:0%
this the right goal? What like why are
we<00:00:30.440><c> doing</c><00:00:30.760><c> this?</c><00:00:31.440><c> Why</c><00:00:31.680><c> can't</c><00:00:31.920><c> we</c><00:00:32.160><c> just</c><00:00:32.520><c> build</c>

00:00:33.150 --> 00:00:33.160 align:start position:0%
we doing this? Why can't we just build
 

00:00:33.160 --> 00:00:35.910 align:start position:0%
we doing this? Why can't we just build
AI<00:00:33.480><c> systems,</c><00:00:34.560><c> which</c><00:00:34.800><c> are</c><00:00:35.040><c> systems</c><00:00:35.600><c> that</c><00:00:35.720><c> have</c>

00:00:35.910 --> 00:00:35.920 align:start position:0%
AI systems, which are systems that have
 

00:00:35.920 --> 00:00:37.950 align:start position:0%
AI systems, which are systems that have
nothing<00:00:36.280><c> to</c><00:00:36.440><c> do</c><00:00:37.160><c> with</c><00:00:37.320><c> the</c><00:00:37.440><c> statistical</c>

00:00:37.950 --> 00:00:37.960 align:start position:0%
nothing to do with the statistical
 

00:00:37.960 --> 00:00:39.510 align:start position:0%
nothing to do with the statistical
engines<00:00:38.280><c> that</c><00:00:38.400><c> they're</c><00:00:38.520><c> trying</c><00:00:38.760><c> to</c><00:00:38.840><c> build</c><00:00:39.280><c> to</c>

00:00:39.510 --> 00:00:39.520 align:start position:0%
engines that they're trying to build to
 

00:00:39.520 --> 00:00:41.350 align:start position:0%
engines that they're trying to build to
duplicate<00:00:40.240><c> the</c><00:00:40.320><c> human</c><00:00:40.560><c> brain?</c>

00:00:41.350 --> 00:00:41.360 align:start position:0%
duplicate the human brain?
 

00:00:41.360 --> 00:00:43.390 align:start position:0%
duplicate the human brain?
&gt;&gt; That<00:00:42.000><c> might</c><00:00:42.200><c> be</c><00:00:42.320><c> the</c><00:00:42.400><c> most</c><00:00:42.680><c> uncomfortable</c>

00:00:43.390 --> 00:00:43.400 align:start position:0%
&gt;&gt; That might be the most uncomfortable
 

00:00:43.400 --> 00:00:45.870 align:start position:0%
&gt;&gt; That might be the most uncomfortable
question<00:00:43.880><c> I've</c><00:00:44.000><c> heard</c><00:00:44.160><c> about</c><00:00:44.560><c> AI</c><00:00:44.960><c> this</c><00:00:45.240><c> year.</c>

00:00:45.870 --> 00:00:45.880 align:start position:0%
question I've heard about AI this year.
 

00:00:45.880 --> 00:00:48.630 align:start position:0%
question I've heard about AI this year.
Not<00:00:46.320><c> can</c><00:00:46.520><c> AI</c><00:00:46.800><c> become</c><00:00:47.120><c> smarter</c><00:00:47.520><c> than</c><00:00:47.680><c> humans,</c>

00:00:48.630 --> 00:00:48.640 align:start position:0%
Not can AI become smarter than humans,
 

00:00:48.640 --> 00:00:50.790 align:start position:0%
Not can AI become smarter than humans,
not<00:00:49.040><c> will</c><00:00:49.200><c> AI</c><00:00:49.360><c> destroy</c><00:00:49.800><c> humanity,</c><00:00:50.640><c> but</c>

00:00:50.790 --> 00:00:50.800 align:start position:0%
not will AI destroy humanity, but
 

00:00:50.800 --> 00:00:52.510 align:start position:0%
not will AI destroy humanity, but
something<00:00:51.160><c> much</c><00:00:51.400><c> simpler.</c><00:00:52.120><c> Why</c><00:00:52.320><c> are</c><00:00:52.400><c> we</c>

00:00:52.510 --> 00:00:52.520 align:start position:0%
something much simpler. Why are we
 

00:00:52.520 --> 00:00:53.990 align:start position:0%
something much simpler. Why are we
trying<00:00:52.840><c> to</c><00:00:52.920><c> build</c><00:00:53.120><c> machines</c><00:00:53.560><c> that</c><00:00:53.680><c> replace</c>

00:00:53.990 --> 00:00:54.000 align:start position:0%
trying to build machines that replace
 

00:00:54.000 --> 00:00:56.430 align:start position:0%
trying to build machines that replace
people<00:00:54.720><c> instead</c><00:00:55.000><c> of</c><00:00:55.080><c> helping</c><00:00:55.400><c> people?</c><00:00:56.240><c> Think</c>

00:00:56.430 --> 00:00:56.440 align:start position:0%
people instead of helping people? Think
 

00:00:56.440 --> 00:00:58.270 align:start position:0%
people instead of helping people? Think
about<00:00:56.680><c> it</c><00:00:56.760><c> for</c><00:00:56.880><c> a</c><00:00:56.920><c> second.</c><00:00:57.640><c> When</c><00:00:57.840><c> cars</c><00:00:58.160><c> were</c>

00:00:58.270 --> 00:00:58.280 align:start position:0%
about it for a second. When cars were
 

00:00:58.280 --> 00:01:00.630 align:start position:0%
about it for a second. When cars were
invented,<00:00:58.960><c> they</c><00:00:59.080><c> helped</c><00:00:59.320><c> us</c><00:00:59.440><c> travel</c><00:00:59.880><c> faster.</c>

00:01:00.630 --> 00:01:00.640 align:start position:0%
invented, they helped us travel faster.
 

00:01:00.640 --> 00:01:02.630 align:start position:0%
invented, they helped us travel faster.
When<00:01:00.840><c> smartphones</c><00:01:01.400><c> became</c><00:01:01.720><c> common,</c><00:01:02.480><c> they</c>

00:01:02.630 --> 00:01:02.640 align:start position:0%
When smartphones became common, they
 

00:01:02.640 --> 00:01:04.630 align:start position:0%
When smartphones became common, they
connected<00:01:03.200><c> us</c><00:01:03.320><c> to</c><00:01:03.440><c> the</c><00:01:03.560><c> whole</c><00:01:04.000><c> world.</c>

00:01:04.630 --> 00:01:04.640 align:start position:0%
connected us to the whole world.
 

00:01:04.640 --> 00:01:06.510 align:start position:0%
connected us to the whole world.
Technology<00:01:05.280><c> has</c><00:01:05.480><c> always</c><00:01:05.840><c> been</c><00:01:06.000><c> about</c><00:01:06.240><c> giving</c>

00:01:06.510 --> 00:01:06.520 align:start position:0%
Technology has always been about giving
 

00:01:06.520 --> 00:01:08.790 align:start position:0%
Technology has always been about giving
humans<00:01:06.920><c> more</c><00:01:07.120><c> power,</c><00:01:08.120><c> but</c><00:01:08.280><c> for</c><00:01:08.360><c> the</c><00:01:08.480><c> first</c>

00:01:08.790 --> 00:01:08.800 align:start position:0%
humans more power, but for the first
 

00:01:08.800 --> 00:01:10.990 align:start position:0%
humans more power, but for the first
time,<00:01:09.400><c> millions</c><00:01:09.840><c> of</c><00:01:09.960><c> people</c><00:01:10.280><c> are</c><00:01:10.480><c> asking</c><00:01:10.920><c> a</c>

00:01:10.990 --> 00:01:11.000 align:start position:0%
time, millions of people are asking a
 

00:01:11.000 --> 00:01:12.910 align:start position:0%
time, millions of people are asking a
completely<00:01:11.520><c> different</c><00:01:11.840><c> question.</c><00:01:12.680><c> What</c>

00:01:12.910 --> 00:01:12.920 align:start position:0%
completely different question. What
 

00:01:12.920 --> 00:01:15.030 align:start position:0%
completely different question. What
happens<00:01:13.480><c> when</c><00:01:13.600><c> technology</c><00:01:14.240><c> starts</c><00:01:14.520><c> replacing</c>

00:01:15.030 --> 00:01:15.040 align:start position:0%
happens when technology starts replacing
 

00:01:15.040 --> 00:01:17.030 align:start position:0%
happens when technology starts replacing
the<00:01:15.120><c> person</c><00:01:15.960><c> instead</c><00:01:16.280><c> of</c><00:01:16.400><c> just</c><00:01:16.640><c> helping</c><00:01:16.960><c> the</c>

00:01:17.030 --> 00:01:17.040 align:start position:0%
the person instead of just helping the
 

00:01:17.040 --> 00:01:19.430 align:start position:0%
the person instead of just helping the
person?<00:01:17.920><c> And</c><00:01:18.080><c> that's</c><00:01:18.280><c> exactly</c><00:01:18.760><c> why</c><00:01:19.000><c> I</c><00:01:19.080><c> wanted</c>

00:01:19.430 --> 00:01:19.440 align:start position:0%
person? And that's exactly why I wanted
 

00:01:19.440 --> 00:01:21.870 align:start position:0%
person? And that's exactly why I wanted
to<00:01:19.560><c> react</c><00:01:19.920><c> to</c><00:01:20.040><c> this</c><00:01:20.240><c> podcast.</c><00:01:21.320><c> A</c><00:01:21.400><c> 2-hour</c>

00:01:21.870 --> 00:01:21.880 align:start position:0%
to react to this podcast. A 2-hour
 

00:01:21.880 --> 00:01:23.950 align:start position:0%
to react to this podcast. A 2-hour
conversation<00:01:22.560><c> between</c><00:01:22.920><c> Steven</c><00:01:23.360><c> Bartlett</c><00:01:23.840><c> and</c>

00:01:23.950 --> 00:01:23.960 align:start position:0%
conversation between Steven Bartlett and
 

00:01:23.960 --> 00:01:26.550 align:start position:0%
conversation between Steven Bartlett and
journalist<00:01:24.520><c> Karen</c><00:01:24.840><c> Hao,</c><00:01:25.720><c> a</c><00:01:25.800><c> woman</c><00:01:26.080><c> who</c><00:01:26.200><c> spent</c>

00:01:26.550 --> 00:01:26.560 align:start position:0%
journalist Karen Hao, a woman who spent
 

00:01:26.560 --> 00:01:29.030 align:start position:0%
journalist Karen Hao, a woman who spent
8<00:01:26.800><c> years</c><00:01:27.160><c> and</c><00:01:27.400><c> over</c><00:01:27.640><c> 300</c><00:01:28.240><c> interviews</c>

00:01:29.030 --> 00:01:29.040 align:start position:0%
8 years and over 300 interviews
 

00:01:29.040 --> 00:01:30.710 align:start position:0%
8 years and over 300 interviews
investigating<00:01:29.720><c> what's</c><00:01:30.000><c> really</c><00:01:30.360><c> happening</c>

00:01:30.710 --> 00:01:30.720 align:start position:0%
investigating what's really happening
 

00:01:30.720 --> 00:01:33.310 align:start position:0%
investigating what's really happening
inside<00:01:31.160><c> companies</c><00:01:31.680><c> like</c><00:01:31.880><c> OpenAI.</c>

00:01:33.310 --> 00:01:33.320 align:start position:0%
inside companies like OpenAI.
 

00:01:33.320 --> 00:01:35.310 align:start position:0%
inside companies like OpenAI.
And<00:01:33.480><c> I</c><00:01:33.520><c> need</c><00:01:33.720><c> to</c><00:01:33.800><c> be</c><00:01:33.920><c> honest</c><00:01:34.200><c> with</c><00:01:34.400><c> you.</c><00:01:35.120><c> Some</c>

00:01:35.310 --> 00:01:35.320 align:start position:0%
And I need to be honest with you. Some
 

00:01:35.320 --> 00:01:37.390 align:start position:0%
And I need to be honest with you. Some
of<00:01:35.440><c> what</c><00:01:35.560><c> she</c><00:01:35.760><c> says</c><00:01:36.080><c> made</c><00:01:36.280><c> me</c><00:01:36.440><c> angry.</c><00:01:37.080><c> Some</c><00:01:37.280><c> of</c>

00:01:37.390 --> 00:01:37.400 align:start position:0%
of what she says made me angry. Some of
 

00:01:37.400 --> 00:01:39.230 align:start position:0%
of what she says made me angry. Some of
it<00:01:37.520><c> made</c><00:01:37.720><c> me</c><00:01:37.840><c> question</c><00:01:38.320><c> things</c><00:01:38.680><c> I</c><00:01:38.760><c> use</c><00:01:39.000><c> every</c>

00:01:39.230 --> 00:01:39.240 align:start position:0%
it made me question things I use every
 

00:01:39.240 --> 00:01:41.870 align:start position:0%
it made me question things I use every
day,<00:01:39.760><c> and</c><00:01:39.920><c> one</c><00:01:40.120><c> part</c><00:01:40.880><c> about</c><00:01:41.120><c> a</c><00:01:41.160><c> small</c><00:01:41.480><c> town</c><00:01:41.760><c> in</c>

00:01:41.870 --> 00:01:41.880 align:start position:0%
day, and one part about a small town in
 

00:01:41.880 --> 00:01:44.870 align:start position:0%
day, and one part about a small town in
Tennessee<00:01:42.920><c> genuinely</c><00:01:43.440><c> disturbed</c><00:01:44.000><c> me.</c>

00:01:44.870 --> 00:01:44.880 align:start position:0%
Tennessee genuinely disturbed me.
 

00:01:44.880 --> 00:01:46.710 align:start position:0%
Tennessee genuinely disturbed me.
Let's<00:01:45.200><c> walk</c><00:01:45.440><c> through</c><00:01:45.680><c> it</c><00:01:46.000><c> and</c><00:01:46.120><c> give</c><00:01:46.240><c> my</c><00:01:46.400><c> honest</c>

00:01:46.710 --> 00:01:46.720 align:start position:0%
Let's walk through it and give my honest
 

00:01:46.720 --> 00:01:48.310 align:start position:0%
Let's walk through it and give my honest
opinion<00:01:47.080><c> about</c><00:01:47.360><c> it.</c>

00:01:48.310 --> 00:01:48.320 align:start position:0%
opinion about it.
 

00:01:48.320 --> 00:01:50.470 align:start position:0%
opinion about it.
I<00:01:48.480><c> want</c><00:01:48.720><c> to</c><00:01:48.800><c> give</c><00:01:48.960><c> you</c><00:01:49.080><c> some</c><00:01:49.280><c> background</c><00:01:49.880><c> first</c>

00:01:50.470 --> 00:01:50.480 align:start position:0%
I want to give you some background first
 

00:01:50.480 --> 00:01:52.350 align:start position:0%
I want to give you some background first
because<00:01:50.800><c> you</c><00:01:50.920><c> can't</c><00:01:51.280><c> understand</c><00:01:51.920><c> the</c><00:01:52.040><c> fight</c>

00:01:52.350 --> 00:01:52.360 align:start position:0%
because you can't understand the fight
 

00:01:52.360 --> 00:01:54.590 align:start position:0%
because you can't understand the fight
without<00:01:52.680><c> understanding</c><00:01:53.400><c> the</c><00:01:53.520><c> field.</c>

00:01:54.590 --> 00:01:54.600 align:start position:0%
without understanding the field.
 

00:01:54.600 --> 00:01:57.350 align:start position:0%
without understanding the field.
&gt;&gt; So,<00:01:54.880><c> this</c><00:01:55.080><c> was</c><00:01:55.240><c> back</c><00:01:55.520><c> in</c><00:01:55.600><c> 1956</c>

00:01:57.350 --> 00:01:57.360 align:start position:0%
&gt;&gt; So, this was back in 1956
 

00:01:57.360 --> 00:01:59.070 align:start position:0%
&gt;&gt; So, this was back in 1956
and<00:01:57.901><c> [music]</c><00:01:58.240><c> there</c><00:01:58.480><c> were</c><00:01:58.720><c> a</c><00:01:58.760><c> group</c><00:01:59.000><c> of</c>

00:01:59.070 --> 00:01:59.080 align:start position:0%
and [music] there were a group of
 

00:01:59.080 --> 00:02:00.350 align:start position:0%
and [music] there were a group of
scientists<00:01:59.560><c> that</c><00:01:59.680><c> gathered</c><00:01:59.960><c> at</c><00:02:00.040><c> Dartmouth</c>

00:02:00.350 --> 00:02:00.360 align:start position:0%
scientists that gathered at Dartmouth
 

00:02:00.360 --> 00:02:02.990 align:start position:0%
scientists that gathered at Dartmouth
University<00:02:01.040><c> to</c><00:02:01.880><c> start</c><00:02:02.200><c> a</c><00:02:02.280><c> new</c><00:02:02.480><c> discipline,</c><00:02:02.960><c> a</c>

00:02:02.990 --> 00:02:03.000 align:start position:0%
University to start a new discipline, a
 

00:02:03.000 --> 00:02:05.030 align:start position:0%
University to start a new discipline, a
scientific<00:02:03.479><c> discipline,</c><00:02:03.960><c> to</c><00:02:04.120><c> try</c><00:02:04.520><c> and</c><00:02:04.720><c> chase</c>

00:02:05.030 --> 00:02:05.040 align:start position:0%
scientific discipline, to try and chase
 

00:02:05.040 --> 00:02:07.950 align:start position:0%
scientific discipline, to try and chase
an<00:02:05.160><c> ambition.</c><00:02:06.280><c> And</c><00:02:06.680><c> specifically,</c><00:02:07.840><c> an</c>

00:02:07.950 --> 00:02:07.960 align:start position:0%
an ambition. And specifically, an
 

00:02:07.960 --> 00:02:09.070 align:start position:0%
an ambition. And specifically, an
assistant<00:02:08.320><c> professor</c><00:02:08.679><c> at</c><00:02:08.759><c> Dartmouth</c>

00:02:09.070 --> 00:02:09.080 align:start position:0%
assistant professor at Dartmouth
 

00:02:09.080 --> 00:02:11.390 align:start position:0%
assistant professor at Dartmouth
University,<00:02:09.640><c> John</c><00:02:09.880><c> McCarthy,</c><00:02:10.880><c> decided</c><00:02:11.280><c> to</c>

00:02:11.390 --> 00:02:11.400 align:start position:0%
University, John McCarthy, decided to
 

00:02:11.400 --> 00:02:12.830 align:start position:0%
University, John McCarthy, decided to
name<00:02:11.680><c> this</c><00:02:11.840><c> discipline</c><00:02:12.280><c> artificial</c>

00:02:12.830 --> 00:02:12.840 align:start position:0%
name this discipline artificial
 

00:02:12.840 --> 00:02:14.550 align:start position:0%
name this discipline artificial
intelligence.

00:02:14.550 --> 00:02:14.560 align:start position:0%
intelligence.
 

00:02:14.560 --> 00:02:16.110 align:start position:0%
intelligence.
This<00:02:14.760><c> was</c><00:02:14.880><c> not</c><00:02:15.080><c> the</c><00:02:15.160><c> first</c><00:02:15.560><c> name</c><00:02:15.840><c> that</c><00:02:16.000><c> he</c>

00:02:16.110 --> 00:02:16.120 align:start position:0%
This was not the first name that he
 

00:02:16.120 --> 00:02:18.310 align:start position:0%
This was not the first name that he
tried.<00:02:16.959><c> The</c><00:02:17.080><c> previous</c><00:02:17.520><c> year,</c><00:02:17.760><c> he</c><00:02:17.880><c> tried</c><00:02:18.200><c> to</c>

00:02:18.310 --> 00:02:18.320 align:start position:0%
tried. The previous year, he tried to
 

00:02:18.320 --> 00:02:21.270 align:start position:0%
tried. The previous year, he tried to
name<00:02:18.560><c> it</c><00:02:18.760><c> Automata</c><00:02:19.360><c> Studies.</c><00:02:20.320><c> And</c><00:02:20.760><c> the</c><00:02:20.880><c> reason</c>

00:02:21.270 --> 00:02:21.280 align:start position:0%
name it Automata Studies. And the reason
 

00:02:21.280 --> 00:02:22.230 align:start position:0%
name it Automata Studies. And the reason
why<00:02:21.440><c> some</c><00:02:21.600><c> of</c><00:02:21.680><c> his</c><00:02:21.800><c> colleagues</c><00:02:22.160><c> were</c>

00:02:22.230 --> 00:02:22.240 align:start position:0%
why some of his colleagues were
 

00:02:22.240 --> 00:02:25.030 align:start position:0%
why some of his colleagues were
concerned<00:02:22.680><c> about</c><00:02:22.880><c> this</c><00:02:23.040><c> name</c><00:02:23.600><c> was</c><00:02:23.920><c> because</c><00:02:24.800><c> it</c>

00:02:25.030 --> 00:02:25.040 align:start position:0%
concerned about this name was because it
 

00:02:25.040 --> 00:02:28.150 align:start position:0%
concerned about this name was because it
pegged<00:02:25.720><c> the</c><00:02:25.880><c> idea</c><00:02:26.280><c> of</c><00:02:26.360><c> this</c><00:02:26.560><c> discipline</c><00:02:27.520><c> to</c>

00:02:28.150 --> 00:02:28.160 align:start position:0%
pegged the idea of this discipline to
 

00:02:28.160 --> 00:02:29.910 align:start position:0%
pegged the idea of this discipline to
recreating<00:02:28.720><c> human</c><00:02:29.000><c> intelligence.</c>

00:02:29.910 --> 00:02:29.920 align:start position:0%
recreating human intelligence.
 

00:02:29.920 --> 00:02:31.390 align:start position:0%
recreating human intelligence.
&gt;&gt; His<00:02:30.080><c> colleagues</c><00:02:30.440><c> were</c><00:02:30.560><c> deeply</c><00:02:30.920><c> concerned</c>

00:02:31.390 --> 00:02:31.400 align:start position:0%
&gt;&gt; His colleagues were deeply concerned
 

00:02:31.400 --> 00:02:33.870 align:start position:0%
&gt;&gt; His colleagues were deeply concerned
about<00:02:31.640><c> the</c><00:02:31.720><c> term</c><00:02:32.280><c> artificial</c><00:02:32.840><c> intelligence</c>

00:02:33.870 --> 00:02:33.880 align:start position:0%
about the term artificial intelligence
 

00:02:33.880 --> 00:02:35.710 align:start position:0%
about the term artificial intelligence
because<00:02:34.160><c> it</c><00:02:34.280><c> tied</c><00:02:34.600><c> the</c><00:02:34.680><c> field</c><00:02:34.959><c> to</c><00:02:35.040><c> the</c><00:02:35.160><c> goal</c><00:02:35.520><c> of</c>

00:02:35.710 --> 00:02:35.720 align:start position:0%
because it tied the field to the goal of
 

00:02:35.720 --> 00:02:38.510 align:start position:0%
because it tied the field to the goal of
recreating<00:02:36.360><c> human</c><00:02:36.680><c> intelligence.</c><00:02:37.959><c> Both</c><00:02:38.200><c> then</c>

00:02:38.510 --> 00:02:38.520 align:start position:0%
recreating human intelligence. Both then
 

00:02:38.520 --> 00:02:40.030 align:start position:0%
recreating human intelligence. Both then
and<00:02:38.640><c> now,</c><00:02:38.959><c> there</c><00:02:39.120><c> is</c><00:02:39.320><c> absolutely</c><00:02:39.840><c> no</c>

00:02:40.030 --> 00:02:40.040 align:start position:0%
and now, there is absolutely no
 

00:02:40.040 --> 00:02:41.830 align:start position:0%
and now, there is absolutely no
scientific<00:02:40.560><c> consensus</c><00:02:41.200><c> on</c><00:02:41.360><c> what</c><00:02:41.560><c> human</c>

00:02:41.830 --> 00:02:41.840 align:start position:0%
scientific consensus on what human
 

00:02:41.840 --> 00:02:44.110 align:start position:0%
scientific consensus on what human
intelligence<00:02:42.400><c> actually</c><00:02:42.800><c> is.</c><00:02:43.760><c> There</c><00:02:43.840><c> is</c><00:02:43.959><c> no</c>

00:02:44.110 --> 00:02:44.120 align:start position:0%
intelligence actually is. There is no
 

00:02:44.120 --> 00:02:45.990 align:start position:0%
intelligence actually is. There is no
single<00:02:44.440><c> definition</c><00:02:45.000><c> from</c><00:02:45.160><c> biology,</c>

00:02:45.990 --> 00:02:46.000 align:start position:0%
single definition from biology,
 

00:02:46.000 --> 00:02:48.870 align:start position:0%
single definition from biology,
psychology,<00:02:46.760><c> or</c><00:02:46.920><c> neurology,</c><00:02:47.880><c> which</c><00:02:48.120><c> means,</c>

00:02:48.870 --> 00:02:48.880 align:start position:0%
psychology, or neurology, which means,
 

00:02:48.880 --> 00:02:51.310 align:start position:0%
psychology, or neurology, which means,
as<00:02:49.040><c> Karen</c><00:02:49.360><c> points</c><00:02:49.680><c> out,</c><00:02:50.160><c> when</c><00:02:50.360><c> today's</c><00:02:50.920><c> AI</c>

00:02:51.310 --> 00:02:51.320 align:start position:0%
as Karen points out, when today's AI
 

00:02:51.320 --> 00:02:54.030 align:start position:0%
as Karen points out, when today's AI
companies<00:02:52.040><c> talk</c><00:02:52.280><c> about</c><00:02:52.520><c> building</c><00:02:53.000><c> AGI,</c>

00:02:54.030 --> 00:02:54.040 align:start position:0%
companies talk about building AGI,
 

00:02:54.040 --> 00:02:56.110 align:start position:0%
companies talk about building AGI,
artificial<00:02:54.560><c> general</c><00:02:54.920><c> intelligence,</c><00:02:55.920><c> they're</c>

00:02:56.110 --> 00:02:56.120 align:start position:0%
artificial general intelligence, they're
 

00:02:56.120 --> 00:02:58.110 align:start position:0%
artificial general intelligence, they're
chasing<00:02:56.520><c> a</c><00:02:56.560><c> finish</c><00:02:56.920><c> line</c><00:02:57.360><c> that</c><00:02:57.600><c> literally</c>

00:02:58.110 --> 00:02:58.120 align:start position:0%
chasing a finish line that literally
 

00:02:58.120 --> 00:03:00.190 align:start position:0%
chasing a finish line that literally
doesn't<00:02:58.480><c> have</c><00:02:58.720><c> fixed</c><00:02:59.040><c> coordinates.</c>

00:03:00.190 --> 00:03:00.200 align:start position:0%
doesn't have fixed coordinates.
 

00:03:00.200 --> 00:03:00.990 align:start position:0%
doesn't have fixed coordinates.
&gt;&gt; Well,

00:03:00.990 --> 00:03:01.000 align:start position:0%
&gt;&gt; Well,
 

00:03:01.000 --> 00:03:04.270 align:start position:0%
&gt;&gt; Well,
there<00:03:01.160><c> are</c><00:03:01.239><c> no</c><00:03:01.440><c> goalposts</c><00:03:02.600><c> for</c><00:03:03.120><c> this</c><00:03:03.400><c> field,</c>

00:03:04.270 --> 00:03:04.280 align:start position:0%
there are no goalposts for this field,
 

00:03:04.280 --> 00:03:05.630 align:start position:0%
there are no goalposts for this field,
and<00:03:04.480><c> there</c><00:03:04.560><c> are</c><00:03:04.600><c> no</c><00:03:04.760><c> goalposts</c><00:03:05.360><c> for</c><00:03:05.480><c> the</c>

00:03:05.630 --> 00:03:05.640 align:start position:0%
and there are no goalposts for the
 

00:03:05.640 --> 00:03:07.270 align:start position:0%
and there are no goalposts for the
industry<00:03:06.120><c> when</c><00:03:06.280><c> they</c><00:03:06.440><c> say</c><00:03:06.840><c> that</c><00:03:07.000><c> they</c><00:03:07.120><c> are</c>

00:03:07.270 --> 00:03:07.280 align:start position:0%
industry when they say that they are
 

00:03:07.280 --> 00:03:10.350 align:start position:0%
industry when they say that they are
ultimately<00:03:07.760><c> trying</c><00:03:08.080><c> to</c><00:03:08.200><c> recreate</c><00:03:09.320><c> AI</c><00:03:09.600><c> systems</c>

00:03:10.350 --> 00:03:10.360 align:start position:0%
ultimately trying to recreate AI systems
 

00:03:10.360 --> 00:03:12.870 align:start position:0%
ultimately trying to recreate AI systems
that<00:03:10.560><c> would</c><00:03:10.720><c> be</c><00:03:10.959><c> as</c><00:03:11.160><c> smart</c><00:03:11.440><c> as</c><00:03:11.560><c> humans.</c><00:03:12.680><c> How</c><00:03:12.800><c> do</c>

00:03:12.870 --> 00:03:12.880 align:start position:0%
that would be as smart as humans. How do
 

00:03:12.880 --> 00:03:15.310 align:start position:0%
that would be as smart as humans. How do
we<00:03:12.959><c> even</c><00:03:13.160><c> define</c><00:03:13.920><c> what</c><00:03:14.040><c> that</c><00:03:14.239><c> means?</c><00:03:14.840><c> And</c><00:03:15.120><c> when</c>

00:03:15.310 --> 00:03:15.320 align:start position:0%
we even define what that means? And when
 

00:03:15.320 --> 00:03:17.390 align:start position:0%
we even define what that means? And when
are<00:03:15.440><c> we</c><00:03:15.760><c> going</c><00:03:16.040><c> to</c><00:03:16.120><c> get</c><00:03:16.280><c> there</c><00:03:16.600><c> if</c><00:03:16.840><c> we</c><00:03:17.040><c> don't</c>

00:03:17.390 --> 00:03:17.400 align:start position:0%
are we going to get there if we don't
 

00:03:17.400 --> 00:03:20.590 align:start position:0%
are we going to get there if we don't
know<00:03:18.160><c> how</c><00:03:18.280><c> to</c><00:03:18.400><c> define</c><00:03:19.239><c> the</c><00:03:19.320><c> destination?</c>

00:03:20.590 --> 00:03:20.600 align:start position:0%
know how to define the destination?
 

00:03:20.600 --> 00:03:24.390 align:start position:0%
know how to define the destination?
&gt;&gt; Karen<00:03:20.920><c> shows</c><00:03:21.239><c> how</c><00:03:21.480><c> OpenAI</c><00:03:22.360><c> has</c><00:03:22.600><c> redefined</c><00:03:23.519><c> AGI</c>

00:03:24.390 --> 00:03:24.400 align:start position:0%
&gt;&gt; Karen shows how OpenAI has redefined AGI
 

00:03:24.400 --> 00:03:26.470 align:start position:0%
&gt;&gt; Karen shows how OpenAI has redefined AGI
depending<00:03:24.720><c> on</c><00:03:24.840><c> who's</c><00:03:25.120><c> in</c><00:03:25.200><c> the</c><00:03:25.320><c> room.</c><00:03:26.000><c> When</c><00:03:26.200><c> Sam</c>

00:03:26.470 --> 00:03:26.480 align:start position:0%
depending on who's in the room. When Sam
 

00:03:26.480 --> 00:03:28.830 align:start position:0%
depending on who's in the room. When Sam
Altman<00:03:26.840><c> talks</c><00:03:27.080><c> to</c><00:03:27.200><c> Congress,</c><00:03:27.959><c> AGI</c><00:03:28.560><c> is</c><00:03:28.720><c> the</c>

00:03:28.830 --> 00:03:28.840 align:start position:0%
Altman talks to Congress, AGI is the
 

00:03:28.840 --> 00:03:31.030 align:start position:0%
Altman talks to Congress, AGI is the
thing<00:03:29.080><c> that</c><00:03:29.280><c> will</c><00:03:29.760><c> cure</c><00:03:30.040><c> cancer,</c><00:03:30.760><c> solve</c>

00:03:31.030 --> 00:03:31.040 align:start position:0%
thing that will cure cancer, solve
 

00:03:31.040 --> 00:03:33.630 align:start position:0%
thing that will cure cancer, solve
climate<00:03:31.400><c> change,</c><00:03:31.959><c> cure</c><00:03:32.200><c> poverty.</c><00:03:33.320><c> When</c><00:03:33.440><c> he's</c>

00:03:33.630 --> 00:03:33.640 align:start position:0%
climate change, cure poverty. When he's
 

00:03:33.640 --> 00:03:35.630 align:start position:0%
climate change, cure poverty. When he's
selling<00:03:33.959><c> to</c><00:03:34.080><c> regular</c><00:03:34.480><c> customers</c><00:03:34.959><c> like</c><00:03:35.200><c> us,</c>

00:03:35.630 --> 00:03:35.640 align:start position:0%
selling to regular customers like us,
 

00:03:35.640 --> 00:03:37.750 align:start position:0%
selling to regular customers like us,
it's<00:03:35.800><c> the</c><00:03:35.920><c> most</c><00:03:36.280><c> amazing</c><00:03:36.920><c> digital</c><00:03:37.280><c> assistant</c>

00:03:37.750 --> 00:03:37.760 align:start position:0%
it's the most amazing digital assistant
 

00:03:37.760 --> 00:03:39.949 align:start position:0%
it's the most amazing digital assistant
you'll<00:03:38.000><c> ever</c><00:03:38.200><c> have.</c><00:03:38.959><c> When</c><00:03:39.200><c> Microsoft</c><00:03:39.800><c> was</c>

00:03:39.949 --> 00:03:39.959 align:start position:0%
you'll ever have. When Microsoft was
 

00:03:39.959 --> 00:03:42.590 align:start position:0%
you'll ever have. When Microsoft was
writing<00:03:40.239><c> the</c><00:03:40.320><c> investment</c><00:03:40.800><c> contract,</c><00:03:41.519><c> AGI</c><00:03:42.440><c> was</c>

00:03:42.590 --> 00:03:42.600 align:start position:0%
writing the investment contract, AGI was
 

00:03:42.600 --> 00:03:44.509 align:start position:0%
writing the investment contract, AGI was
legally<00:03:42.959><c> defined</c><00:03:43.440><c> as</c><00:03:43.519><c> a</c><00:03:43.560><c> system</c><00:03:44.360><c> that</c>

00:03:44.509 --> 00:03:44.519 align:start position:0%
legally defined as a system that
 

00:03:44.519 --> 00:03:47.990 align:start position:0%
legally defined as a system that
generates<00:03:45.000><c> $100</c><00:03:45.600><c> billion</c><00:03:46.400><c> in</c><00:03:46.560><c> revenue.</c>

00:03:47.990 --> 00:03:48.000 align:start position:0%
generates $100 billion in revenue.
 

00:03:48.000 --> 00:03:49.990 align:start position:0%
generates $100 billion in revenue.
&gt;&gt; They<00:03:48.080><c> can</c><00:03:48.239><c> use</c><00:03:48.440><c> it</c><00:03:48.840><c> however</c><00:03:49.200><c> they</c><00:03:49.360><c> want</c><00:03:49.680><c> to,</c>

00:03:49.990 --> 00:03:50.000 align:start position:0%
&gt;&gt; They can use it however they want to,
 

00:03:50.000 --> 00:03:52.150 align:start position:0%
&gt;&gt; They can use it however they want to,
and<00:03:50.160><c> they</c><00:03:50.480><c> can</c><00:03:50.720><c> define</c><00:03:51.120><c> and</c><00:03:51.280><c> redefine</c><00:03:51.800><c> it</c>

00:03:52.150 --> 00:03:52.160 align:start position:0%
and they can define and redefine it
 

00:03:52.160 --> 00:03:53.670 align:start position:0%
and they can define and redefine it
based<00:03:52.480><c> on</c><00:03:52.600><c> what</c><00:03:52.720><c> is</c><00:03:52.840><c> convenient</c><00:03:53.280><c> for</c><00:03:53.400><c> them.</c>

00:03:53.670 --> 00:03:53.680 align:start position:0%
based on what is convenient for them.
 

00:03:53.680 --> 00:03:56.070 align:start position:0%
based on what is convenient for them.
So,<00:03:53.960><c> in</c><00:03:54.160><c> OpenAI's</c><00:03:54.680><c> history,</c><00:03:55.080><c> it</c><00:03:55.240><c> has</c><00:03:55.720><c> defined</c>

00:03:56.070 --> 00:03:56.080 align:start position:0%
So, in OpenAI's history, it has defined
 

00:03:56.080 --> 00:03:58.190 align:start position:0%
So, in OpenAI's history, it has defined
and<00:03:56.200><c> redefined</c><00:03:56.560><c> it</c><00:03:56.600><c> many</c><00:03:56.800><c> times.</c><00:03:57.760><c> When</c><00:03:57.920><c> Sam</c>

00:03:58.190 --> 00:03:58.200 align:start position:0%
and redefined it many times. When Sam
 

00:03:58.200 --> 00:04:00.830 align:start position:0%
and redefined it many times. When Sam
Altman<00:03:58.480><c> is</c><00:03:58.560><c> talking</c><00:03:58.840><c> with</c><00:03:59.000><c> Congress,</c><00:03:59.920><c> AGI</c><00:04:00.640><c> is</c>

00:04:00.830 --> 00:04:00.840 align:start position:0%
Altman is talking with Congress, AGI is
 

00:04:00.840 --> 00:04:02.710 align:start position:0%
Altman is talking with Congress, AGI is
a<00:04:00.880><c> system</c><00:04:01.240><c> that's</c><00:04:01.440><c> going</c><00:04:01.560><c> to</c><00:04:01.600><c> cure</c><00:04:01.800><c> cancer,</c>

00:04:02.710 --> 00:04:02.720 align:start position:0%
a system that's going to cure cancer,
 

00:04:02.720 --> 00:04:05.830 align:start position:0%
a system that's going to cure cancer,
solve<00:04:03.320><c> climate</c><00:04:03.680><c> change,</c><00:04:04.160><c> cure</c><00:04:04.360><c> poverty.</c><00:04:05.480><c> When</c>

00:04:05.830 --> 00:04:05.840 align:start position:0%
solve climate change, cure poverty. When
 

00:04:05.840 --> 00:04:07.670 align:start position:0%
solve climate change, cure poverty. When
he's<00:04:06.000><c> talking</c><00:04:06.400><c> with</c><00:04:06.640><c> consumers</c><00:04:07.440><c> that</c><00:04:07.560><c> he's</c>

00:04:07.670 --> 00:04:07.680 align:start position:0%
he's talking with consumers that he's
 

00:04:07.680 --> 00:04:10.430 align:start position:0%
he's talking with consumers that he's
trying<00:04:07.840><c> to</c><00:04:07.920><c> sell</c><00:04:08.480><c> his</c><00:04:08.680><c> products</c><00:04:09.080><c> to,</c><00:04:09.720><c> it's</c><00:04:10.320><c> the</c>

00:04:10.430 --> 00:04:10.440 align:start position:0%
trying to sell his products to, it's the
 

00:04:10.440 --> 00:04:12.190 align:start position:0%
trying to sell his products to, it's the
most<00:04:10.680><c> amazing</c><00:04:11.120><c> digital</c><00:04:11.440><c> assistant</c><00:04:12.080><c> that</c>

00:04:12.190 --> 00:04:12.200 align:start position:0%
most amazing digital assistant that
 

00:04:12.200 --> 00:04:14.350 align:start position:0%
most amazing digital assistant that
you're<00:04:12.320><c> going</c><00:04:12.440><c> to</c><00:04:12.520><c> ever</c><00:04:12.680><c> going</c><00:04:12.800><c> to</c><00:04:12.880><c> have.</c><00:04:13.880><c> When</c>

00:04:14.350 --> 00:04:14.360 align:start position:0%
you're going to ever going to have. When
 

00:04:14.360 --> 00:04:16.390 align:start position:0%
you're going to ever going to have. When
he<00:04:14.880><c> was</c><00:04:15.040><c> talking</c><00:04:15.360><c> with</c><00:04:15.520><c> Microsoft,</c><00:04:16.120><c> you</c><00:04:16.200><c> know,</c>

00:04:16.390 --> 00:04:16.400 align:start position:0%
he was talking with Microsoft, you know,
 

00:04:16.400 --> 00:04:18.590 align:start position:0%
he was talking with Microsoft, you know,
in<00:04:16.640><c> the</c><00:04:17.000><c> deal</c><00:04:17.359><c> that</c><00:04:17.560><c> OpenAI</c><00:04:17.880><c> and</c><00:04:17.959><c> Microsoft</c>

00:04:18.590 --> 00:04:18.600 align:start position:0%
in the deal that OpenAI and Microsoft
 

00:04:18.600 --> 00:04:20.870 align:start position:0%
in the deal that OpenAI and Microsoft
struck<00:04:19.519><c> where</c><00:04:19.720><c> Microsoft</c><00:04:20.160><c> invested</c><00:04:20.720><c> in</c><00:04:20.799><c> the</c>

00:04:20.870 --> 00:04:20.880 align:start position:0%
struck where Microsoft invested in the
 

00:04:20.880 --> 00:04:23.790 align:start position:0%
struck where Microsoft invested in the
company,<00:04:21.920><c> it</c><00:04:22.120><c> was</c><00:04:22.280><c> defined</c><00:04:23.040><c> as</c><00:04:23.200><c> a</c><00:04:23.240><c> system</c><00:04:23.680><c> that</c>

00:04:23.790 --> 00:04:23.800 align:start position:0%
company, it was defined as a system that
 

00:04:23.800 --> 00:04:25.470 align:start position:0%
company, it was defined as a system that
will<00:04:23.919><c> generate</c><00:04:24.280><c> a</c><00:04:24.360><c> hundred</c><00:04:24.720><c> billion</c><00:04:25.040><c> dollars</c>

00:04:25.470 --> 00:04:25.480 align:start position:0%
will generate a hundred billion dollars
 

00:04:25.480 --> 00:04:28.350 align:start position:0%
will generate a hundred billion dollars
of<00:04:25.640><c> revenue.</c><00:04:26.680><c> And</c><00:04:26.919><c> on</c><00:04:27.080><c> OpenAI's</c><00:04:27.640><c> own</c><00:04:27.919><c> website,</c>

00:04:28.350 --> 00:04:28.360 align:start position:0%
of revenue. And on OpenAI's own website,
 

00:04:28.360 --> 00:04:30.190 align:start position:0%
of revenue. And on OpenAI's own website,
they<00:04:28.440><c> define</c><00:04:28.919><c> it</c><00:04:29.160><c> as</c><00:04:29.360><c> highly</c><00:04:29.720><c> autonomous</c>

00:04:30.190 --> 00:04:30.200 align:start position:0%
they define it as highly autonomous
 

00:04:30.200 --> 00:04:33.550 align:start position:0%
they define it as highly autonomous
systems<00:04:31.320><c> that</c><00:04:31.520><c> outperform</c><00:04:32.080><c> humans</c><00:04:32.640><c> in</c><00:04:32.960><c> most</c>

00:04:33.550 --> 00:04:33.560 align:start position:0%
systems that outperform humans in most
 

00:04:33.560 --> 00:04:36.630 align:start position:0%
systems that outperform humans in most
economically<00:04:34.600><c> valuable</c><00:04:35.520><c> work.</c><00:04:36.160><c> This</c><00:04:36.320><c> is</c><00:04:36.440><c> like</c>

00:04:36.630 --> 00:04:36.640 align:start position:0%
economically valuable work. This is like
 

00:04:36.640 --> 00:04:39.750 align:start position:0%
economically valuable work. This is like
not<00:04:37.000><c> a</c><00:04:37.080><c> coherent</c><00:04:38.360><c> vision</c><00:04:38.919><c> of</c><00:04:39.120><c> one</c><00:04:39.280><c> technology.</c>

00:04:39.750 --> 00:04:39.760 align:start position:0%
not a coherent vision of one technology.
 

00:04:39.760 --> 00:04:42.030 align:start position:0%
not a coherent vision of one technology.
These<00:04:40.000><c> are</c><00:04:40.120><c> very</c><00:04:40.480><c> different</c><00:04:40.760><c> definitions</c>

00:04:42.030 --> 00:04:42.040 align:start position:0%
These are very different definitions
 

00:04:42.040 --> 00:04:44.710 align:start position:0%
These are very different definitions
that<00:04:42.240><c> are</c><00:04:42.360><c> spoken</c><00:04:42.840><c> out</c><00:04:43.040><c> loud</c><00:04:43.440><c> to</c><00:04:43.720><c> the</c><00:04:43.960><c> audience</c>

00:04:44.710 --> 00:04:44.720 align:start position:0%
that are spoken out loud to the audience
 

00:04:44.720 --> 00:04:47.990 align:start position:0%
that are spoken out loud to the audience
that<00:04:44.919><c> needs</c><00:04:45.200><c> to</c><00:04:45.280><c> be</c><00:04:45.400><c> mobilized</c><00:04:46.600><c> to</c><00:04:47.480><c> ward</c><00:04:47.760><c> off</c>

00:04:47.990 --> 00:04:48.000 align:start position:0%
that needs to be mobilized to ward off
 

00:04:48.000 --> 00:04:50.710 align:start position:0%
that needs to be mobilized to ward off
regulation<00:04:48.880><c> or</c><00:04:49.160><c> get</c><00:04:49.440><c> more</c><00:04:49.640><c> consumer</c><00:04:50.080><c> buy-in</c>

00:04:50.710 --> 00:04:50.720 align:start position:0%
regulation or get more consumer buy-in
 

00:04:50.720 --> 00:04:53.190 align:start position:0%
regulation or get more consumer buy-in
into<00:04:51.200><c> the</c><00:04:51.520><c> the</c><00:04:51.680><c> industry's</c><00:04:52.200><c> quest.</c>

00:04:53.190 --> 00:04:53.200 align:start position:0%
into the the industry's quest.
 

00:04:53.200 --> 00:04:55.510 align:start position:0%
into the the industry's quest.
&gt;&gt; Read<00:04:53.440><c> that</c><00:04:53.640><c> again.</c><00:04:54.320><c> Same</c><00:04:54.600><c> three</c><00:04:54.880><c> letters,</c>

00:04:55.510 --> 00:04:55.520 align:start position:0%
&gt;&gt; Read that again. Same three letters,
 

00:04:55.520 --> 00:04:57.350 align:start position:0%
&gt;&gt; Read that again. Same three letters,
completely<00:04:56.080><c> different</c><00:04:56.440><c> meaning</c><00:04:56.919><c> depending</c>

00:04:57.350 --> 00:04:57.360 align:start position:0%
completely different meaning depending
 

00:04:57.360 --> 00:04:59.430 align:start position:0%
completely different meaning depending
on<00:04:57.480><c> who</c><00:04:57.720><c> needs</c><00:04:57.960><c> convincing.</c><00:04:58.960><c> I</c><00:04:59.040><c> don't</c><00:04:59.280><c> think</c>

00:04:59.430 --> 00:04:59.440 align:start position:0%
on who needs convincing. I don't think
 

00:04:59.440 --> 00:05:01.150 align:start position:0%
on who needs convincing. I don't think
that's<00:04:59.640><c> a</c><00:04:59.680><c> coincidence.</c><00:05:00.640><c> I</c><00:05:00.760><c> think</c><00:05:00.919><c> that's</c><00:05:01.120><c> a</c>

00:05:01.150 --> 00:05:01.160 align:start position:0%
that's a coincidence. I think that's a
 

00:05:01.160 --> 00:05:02.670 align:start position:0%
that's a coincidence. I think that's a
strategy.

00:05:02.670 --> 00:05:02.680 align:start position:0%
strategy.
 

00:05:02.680 --> 00:05:04.150 align:start position:0%
strategy.
Here's<00:05:02.880><c> the</c><00:05:03.000><c> clip</c><00:05:03.200><c> that</c><00:05:03.360><c> genuinely</c><00:05:03.840><c> made</c><00:05:04.040><c> me</c>

00:05:04.150 --> 00:05:04.160 align:start position:0%
Here's the clip that genuinely made me
 

00:05:04.160 --> 00:05:05.470 align:start position:0%
Here's the clip that genuinely made me
sit<00:05:04.360><c> up.</c>

00:05:05.470 --> 00:05:05.480 align:start position:0%
sit up.
 

00:05:05.480 --> 00:05:07.190 align:start position:0%
sit up.
&gt;&gt; In<00:05:05.680><c> 2015,</c>

00:05:07.190 --> 00:05:07.200 align:start position:0%
&gt;&gt; In 2015,
 

00:05:07.200 --> 00:05:08.790 align:start position:0%
&gt;&gt; In 2015,
in<00:05:07.280><c> a</c><00:05:07.320><c> blog</c><00:05:07.560><c> post</c><00:05:07.800><c> that</c><00:05:07.919><c> Sam</c><00:05:08.200><c> Altman</c><00:05:08.560><c> wrote</c>

00:05:08.790 --> 00:05:08.800 align:start position:0%
in a blog post that Sam Altman wrote
 

00:05:08.800 --> 00:05:11.030 align:start position:0%
in a blog post that Sam Altman wrote
before<00:05:09.240><c> OpenAI</c><00:05:09.560><c> was</c><00:05:09.720><c> officially</c><00:05:10.400><c> announced,</c>

00:05:11.030 --> 00:05:11.040 align:start position:0%
before OpenAI was officially announced,
 

00:05:11.040 --> 00:05:13.430 align:start position:0%
before OpenAI was officially announced,
he<00:05:11.360><c> explicitly</c><00:05:12.120><c> outlined</c><00:05:12.640><c> the</c><00:05:12.760><c> existential</c>

00:05:13.430 --> 00:05:13.440 align:start position:0%
he explicitly outlined the existential
 

00:05:13.440 --> 00:05:15.270 align:start position:0%
he explicitly outlined the existential
risk<00:05:14.080><c> by</c><00:05:14.200><c> saying,</c><00:05:14.680><c> "Development</c><00:05:15.160><c> of</c>

00:05:15.270 --> 00:05:15.280 align:start position:0%
risk by saying, "Development of
 

00:05:15.280 --> 00:05:17.710 align:start position:0%
risk by saying, "Development of
superhuman<00:05:16.560><c> machine</c><00:05:17.000><c> intelligence</c><00:05:17.600><c> is</c>

00:05:17.710 --> 00:05:17.720 align:start position:0%
superhuman machine intelligence is
 

00:05:17.720 --> 00:05:18.950 align:start position:0%
superhuman machine intelligence is
probably<00:05:18.040><c> the</c><00:05:18.120><c> greatest</c><00:05:18.480><c> threat</c><00:05:18.840><c> to</c><00:05:18.880><c> the</c>

00:05:18.950 --> 00:05:18.960 align:start position:0%
probably the greatest threat to the
 

00:05:18.960 --> 00:05:21.190 align:start position:0%
probably the greatest threat to the
continued<00:05:19.480><c> existence</c><00:05:20.040><c> of</c><00:05:20.120><c> humanity.</c><00:05:21.000><c> There</c>

00:05:21.190 --> 00:05:21.200 align:start position:0%
continued existence of humanity. There
 

00:05:21.200 --> 00:05:22.630 align:start position:0%
continued existence of humanity. There
are<00:05:21.360><c> other</c><00:05:21.560><c> threats</c><00:05:21.960><c> that</c><00:05:22.080><c> I</c><00:05:22.120><c> think</c><00:05:22.360><c> are</c><00:05:22.440><c> more</c>

00:05:22.630 --> 00:05:22.640 align:start position:0%
are other threats that I think are more
 

00:05:22.640 --> 00:05:24.390 align:start position:0%
are other threats that I think are more
certain<00:05:23.000><c> to</c><00:05:23.080><c> happen,</c><00:05:23.480><c> for</c><00:05:23.560><c> example,</c><00:05:24.320><c> an</c>

00:05:24.390 --> 00:05:24.400 align:start position:0%
certain to happen, for example, an
 

00:05:24.400 --> 00:05:25.830 align:start position:0%
certain to happen, for example, an
engineered<00:05:24.800><c> virus,</c>

00:05:25.830 --> 00:05:25.840 align:start position:0%
engineered virus,
 

00:05:25.840 --> 00:05:28.550 align:start position:0%
engineered virus,
but<00:05:26.040><c> AI</c><00:05:26.320><c> is</c><00:05:26.480><c> probably</c><00:05:27.200><c> the</c><00:05:27.280><c> most</c><00:05:27.840><c> likely</c><00:05:28.280><c> way</c>

00:05:28.550 --> 00:05:28.560 align:start position:0%
but AI is probably the most likely way
 

00:05:28.560 --> 00:05:30.150 align:start position:0%
but AI is probably the most likely way
to<00:05:28.720><c> destroy</c><00:05:29.200><c> everything."</c>

00:05:30.150 --> 00:05:30.160 align:start position:0%
to destroy everything."
 

00:05:30.160 --> 00:05:33.270 align:start position:0%
to destroy everything."
&gt;&gt; Back<00:05:30.400><c> in</c><00:05:30.520><c> 2015,</c><00:05:31.520><c> before</c><00:05:32.120><c> OpenAI</c><00:05:32.720><c> officially</c>

00:05:33.270 --> 00:05:33.280 align:start position:0%
&gt;&gt; Back in 2015, before OpenAI officially
 

00:05:33.280 --> 00:05:35.590 align:start position:0%
&gt;&gt; Back in 2015, before OpenAI officially
existed,<00:05:34.040><c> Sam</c><00:05:34.360><c> Altman</c><00:05:34.720><c> wrote</c><00:05:34.880><c> a</c><00:05:34.960><c> blog</c><00:05:35.280><c> post</c>

00:05:35.590 --> 00:05:35.600 align:start position:0%
existed, Sam Altman wrote a blog post
 

00:05:35.600 --> 00:05:37.150 align:start position:0%
existed, Sam Altman wrote a blog post
that<00:05:35.760><c> said,</c><00:05:36.240><c> and</c><00:05:36.360><c> I'm</c><00:05:36.480><c> paraphrasing</c>

00:05:37.150 --> 00:05:37.160 align:start position:0%
that said, and I'm paraphrasing
 

00:05:37.160 --> 00:05:40.030 align:start position:0%
that said, and I'm paraphrasing
carefully<00:05:37.640><c> here,</c><00:05:38.440><c> that</c><00:05:38.600><c> a</c><00:05:38.640><c> superhuman</c><00:05:39.360><c> AI</c><00:05:39.840><c> was</c>

00:05:40.030 --> 00:05:40.040 align:start position:0%
carefully here, that a superhuman AI was
 

00:05:40.040 --> 00:05:41.909 align:start position:0%
carefully here, that a superhuman AI was
probably<00:05:40.440><c> the</c><00:05:40.560><c> single</c><00:05:41.080><c> biggest</c><00:05:41.520><c> threat</c><00:05:41.840><c> to</c>

00:05:41.909 --> 00:05:41.919 align:start position:0%
probably the single biggest threat to
 

00:05:41.919 --> 00:05:44.350 align:start position:0%
probably the single biggest threat to
humanity's<00:05:42.520><c> survival,</c><00:05:43.440><c> bigger</c><00:05:43.760><c> than</c><00:05:43.960><c> almost</c>

00:05:44.350 --> 00:05:44.360 align:start position:0%
humanity's survival, bigger than almost
 

00:05:44.360 --> 00:05:45.710 align:start position:0%
humanity's survival, bigger than almost
anything<00:05:44.720><c> else.</c>

00:05:45.710 --> 00:05:45.720 align:start position:0%
anything else.
 

00:05:45.720 --> 00:05:47.510 align:start position:0%
anything else.
Karen's<00:05:46.160><c> reporting</c><00:05:46.640><c> shows</c><00:05:46.960><c> that</c><00:05:47.120><c> Altman</c>

00:05:47.510 --> 00:05:47.520 align:start position:0%
Karen's reporting shows that Altman
 

00:05:47.520 --> 00:05:49.550 align:start position:0%
Karen's reporting shows that Altman
wrote<00:05:47.760><c> that</c><00:05:47.960><c> specific</c><00:05:48.600><c> line</c><00:05:49.040><c> right</c><00:05:49.280><c> around</c>

00:05:49.550 --> 00:05:49.560 align:start position:0%
wrote that specific line right around
 

00:05:49.560 --> 00:05:51.390 align:start position:0%
wrote that specific line right around
the<00:05:49.680><c> time</c><00:05:49.960><c> he</c><00:05:50.040><c> was</c><00:05:50.160><c> trying</c><00:05:50.440><c> to</c><00:05:50.520><c> convince</c><00:05:51.000><c> Elon</c>

00:05:51.390 --> 00:05:51.400 align:start position:0%
the time he was trying to convince Elon
 

00:05:51.400 --> 00:05:54.510 align:start position:0%
the time he was trying to convince Elon
Musk<00:05:51.960><c> to</c><00:05:52.040><c> co-found</c><00:05:52.640><c> OpenAI</c><00:05:53.400><c> with</c><00:05:53.600><c> him.</c><00:05:54.360><c> And</c>

00:05:54.510 --> 00:05:54.520 align:start position:0%
Musk to co-found OpenAI with him. And
 

00:05:54.520 --> 00:05:57.150 align:start position:0%
Musk to co-found OpenAI with him. And
Musk,<00:05:55.080><c> at</c><00:05:55.200><c> that</c><00:05:55.320><c> exact</c><00:05:55.800><c> same</c><00:05:56.040><c> time,</c><00:05:56.760><c> was</c><00:05:56.920><c> doing</c>

00:05:57.150 --> 00:05:57.160 align:start position:0%
Musk, at that exact same time, was doing
 

00:05:57.160 --> 00:05:59.950 align:start position:0%
Musk, at that exact same time, was doing
famous<00:05:57.600><c> talks,</c><00:05:58.160><c> including</c><00:05:58.680><c> one</c><00:05:58.840><c> at</c><00:05:58.960><c> MIT,</c>

00:05:59.950 --> 00:05:59.960 align:start position:0%
famous talks, including one at MIT,
 

00:05:59.960 --> 00:06:02.150 align:start position:0%
famous talks, including one at MIT,
where<00:06:00.080><c> he</c><00:06:00.200><c> compared</c><00:06:00.640><c> building</c><00:06:00.960><c> advanced</c><00:06:01.480><c> AI</c>

00:06:02.150 --> 00:06:02.160 align:start position:0%
where he compared building advanced AI
 

00:06:02.160 --> 00:06:04.230 align:start position:0%
where he compared building advanced AI
to<00:06:02.400><c> summoning</c><00:06:02.960><c> the</c><00:06:03.080><c> demon.</c>

00:06:04.230 --> 00:06:04.240 align:start position:0%
to summoning the demon.
 

00:06:04.240 --> 00:06:06.470 align:start position:0%
to summoning the demon.
&gt;&gt; And<00:06:04.440><c> so,</c><00:06:04.760><c> in</c><00:06:05.000><c> that</c><00:06:05.200><c> blog</c><00:06:05.440><c> post,</c><00:06:05.760><c> if</c><00:06:05.920><c> you</c><00:06:06.240><c> look</c>

00:06:06.470 --> 00:06:06.480 align:start position:0%
&gt;&gt; And so, in that blog post, if you look
 

00:06:06.480 --> 00:06:08.430 align:start position:0%
&gt;&gt; And so, in that blog post, if you look
at<00:06:06.600><c> the</c><00:06:06.920><c> the</c><00:06:07.040><c> language</c><00:06:07.560><c> that</c><00:06:07.720><c> Altman</c><00:06:08.000><c> uses</c>

00:06:08.430 --> 00:06:08.440 align:start position:0%
at the the language that Altman uses
 

00:06:08.440 --> 00:06:10.350 align:start position:0%
at the the language that Altman uses
side<00:06:08.720><c> by</c><00:06:08.840><c> side</c><00:06:09.280><c> with</c><00:06:09.360><c> the</c><00:06:09.480><c> language</c><00:06:09.840><c> that</c><00:06:10.040><c> Musk</c>

00:06:10.350 --> 00:06:10.360 align:start position:0%
side by side with the language that Musk
 

00:06:10.360 --> 00:06:13.190 align:start position:0%
side by side with the language that Musk
was<00:06:10.480><c> using</c><00:06:10.720><c> at</c><00:06:10.800><c> the</c><00:06:10.880><c> time,</c><00:06:11.800><c> it</c><00:06:12.000><c> mirrors</c><00:06:13.120><c> the</c>

00:06:13.190 --> 00:06:13.200 align:start position:0%
was using at the time, it mirrors the
 

00:06:13.200 --> 00:06:14.990 align:start position:0%
was using at the time, it mirrors the
all<00:06:13.360><c> the</c><00:06:13.440><c> things</c><00:06:13.680><c> that</c><00:06:13.760><c> Musk</c><00:06:13.960><c> was</c><00:06:14.080><c> saying.</c>

00:06:14.990 --> 00:06:15.000 align:start position:0%
all the things that Musk was saying.
 

00:06:15.000 --> 00:06:15.670 align:start position:0%
all the things that Musk was saying.
&gt;&gt; Yeah.

00:06:15.670 --> 00:06:15.680 align:start position:0%
&gt;&gt; Yeah.
 

00:06:15.680 --> 00:06:17.950 align:start position:0%
&gt;&gt; Yeah.
&gt;&gt; years<00:06:15.800><c> ago</c><00:06:15.960><c> Musk</c><00:06:16.280><c> was</c><00:06:16.520><c> going</c><00:06:16.800><c> on</c><00:06:17.440><c> podcasts</c>

00:06:17.950 --> 00:06:17.960 align:start position:0%
&gt;&gt; years ago Musk was going on podcasts
 

00:06:17.960 --> 00:06:19.990 align:start position:0%
&gt;&gt; years ago Musk was going on podcasts
saying,<00:06:18.200><c> tweeting,</c><00:06:18.560><c> whatever,</c><00:06:19.400><c> that</c><00:06:19.919><c> the</c>

00:06:19.990 --> 00:06:20.000 align:start position:0%
saying, tweeting, whatever, that the
 

00:06:20.000 --> 00:06:21.630 align:start position:0%
saying, tweeting, whatever, that the
greatest<00:06:20.480><c> existential</c><00:06:21.000><c> risk</c><00:06:21.120><c> to</c><00:06:21.200><c> humanity</c>

00:06:21.630 --> 00:06:21.640 align:start position:0%
greatest existential risk to humanity
 

00:06:21.640 --> 00:06:22.430 align:start position:0%
greatest existential risk to humanity
was<00:06:21.800><c> AI.</c>

00:06:22.430 --> 00:06:22.440 align:start position:0%
was AI.
 

00:06:22.440 --> 00:06:25.270 align:start position:0%
was AI.
&gt;&gt; Musk<00:06:22.840><c> did</c><00:06:23.080><c> end</c><00:06:23.200><c> up</c><00:06:23.360><c> co-founding</c><00:06:23.960><c> OpenAI.</c><00:06:25.160><c> He</c>

00:06:25.270 --> 00:06:25.280 align:start position:0%
&gt;&gt; Musk did end up co-founding OpenAI. He
 

00:06:25.280 --> 00:06:27.190 align:start position:0%
&gt;&gt; Musk did end up co-founding OpenAI. He
also<00:06:25.560><c> ended</c><00:06:25.800><c> up</c><00:06:25.919><c> leaving,</c><00:06:26.640><c> and</c><00:06:26.760><c> by</c><00:06:26.880><c> his</c><00:06:27.040><c> own</c>

00:06:27.190 --> 00:06:27.200 align:start position:0%
also ended up leaving, and by his own
 

00:06:27.200 --> 00:06:29.350 align:start position:0%
also ended up leaving, and by his own
account,<00:06:27.600><c> feeling</c><00:06:27.919><c> used.</c><00:06:28.720><c> The</c><00:06:28.840><c> two</c><00:06:29.040><c> are</c><00:06:29.160><c> now</c>

00:06:29.350 --> 00:06:29.360 align:start position:0%
account, feeling used. The two are now
 

00:06:29.360 --> 00:06:31.590 align:start position:0%
account, feeling used. The two are now
suing<00:06:29.720><c> each</c><00:06:29.880><c> other.</c>

00:06:31.590 --> 00:06:31.600 align:start position:0%
suing each other.
 

00:06:31.600 --> 00:06:33.350 align:start position:0%
suing each other.
But<00:06:31.720><c> this</c><00:06:31.880><c> next</c><00:06:32.120><c> part</c><00:06:32.760><c> is</c><00:06:32.880><c> where</c><00:06:33.000><c> Karen's</c>

00:06:33.350 --> 00:06:33.360 align:start position:0%
But this next part is where Karen's
 

00:06:33.360 --> 00:06:35.150 align:start position:0%
But this next part is where Karen's
reporting<00:06:33.760><c> feels</c><00:06:34.440><c> like</c><00:06:34.600><c> something</c><00:06:34.880><c> you</c><00:06:35.000><c> need</c>

00:06:35.150 --> 00:06:35.160 align:start position:0%
reporting feels like something you need
 

00:06:35.160 --> 00:06:36.790 align:start position:0%
reporting feels like something you need
to<00:06:35.280><c> actually</c><00:06:35.640><c> know</c><00:06:36.080><c> before</c><00:06:36.360><c> you</c><00:06:36.480><c> open</c><00:06:36.680><c> your</c>

00:06:36.790 --> 00:06:36.800 align:start position:0%
to actually know before you open your
 

00:06:36.800 --> 00:06:38.110 align:start position:0%
to actually know before you open your
chatbot<00:06:37.200><c> app</c><00:06:37.320><c> again.</c>

00:06:38.110 --> 00:06:38.120 align:start position:0%
chatbot app again.
 

00:06:38.120 --> 00:06:39.550 align:start position:0%
chatbot app again.
&gt;&gt; And<00:06:38.280><c> that</c><00:06:38.440><c> is</c><00:06:38.560><c> why</c><00:06:38.720><c> I</c><00:06:38.800><c> call</c><00:06:39.000><c> these</c><00:06:39.120><c> companies</c>

00:06:39.550 --> 00:06:39.560 align:start position:0%
&gt;&gt; And that is why I call these companies
 

00:06:39.560 --> 00:06:40.670 align:start position:0%
&gt;&gt; And that is why I call these companies
empires<00:06:39.960><c> of</c><00:06:40.080><c> AI.</c>

00:06:40.670 --> 00:06:40.680 align:start position:0%
empires of AI.
 

00:06:40.680 --> 00:06:42.990 align:start position:0%
empires of AI.
&gt;&gt; What<00:06:40.800><c> do</c><00:06:40.880><c> you</c><00:06:40.919><c> mean</c><00:06:41.080><c> by</c><00:06:41.200><c> an</c><00:06:41.280><c> imperial</c><00:06:41.720><c> agenda?</c>

00:06:42.990 --> 00:06:43.000 align:start position:0%
&gt;&gt; What do you mean by an imperial agenda?
 

00:06:43.000 --> 00:06:44.270 align:start position:0%
&gt;&gt; What do you mean by an imperial agenda?
What<00:06:43.160><c> What</c><00:06:43.240><c> does</c><00:06:43.360><c> that</c><00:06:43.480><c> term</c><00:06:43.680><c> mean?</c>

00:06:44.270 --> 00:06:44.280 align:start position:0%
What What does that term mean?
 

00:06:44.280 --> 00:06:47.404 align:start position:0%
What What does that term mean?
&gt;&gt; Empire<00:06:44.840><c> is</c><00:06:45.040><c> the</c><00:06:45.320><c> only</c><00:06:46.080><c> metaphor</c><00:06:47.160><c> that</c><00:06:47.360><c> I've</c>

00:06:47.404 --> 00:06:47.414 align:start position:0%
&gt;&gt; Empire is the only metaphor that I've
 

00:06:47.414 --> 00:06:50.110 align:start position:0%
&gt;&gt; Empire is the only metaphor that I've
[music]<00:06:47.640><c> ever</c><00:06:47.960><c> found</c><00:06:48.640><c> to</c><00:06:48.960><c> fully</c><00:06:49.400><c> encapsulate</c>

00:06:50.110 --> 00:06:50.120 align:start position:0%
[music] ever found to fully encapsulate
 

00:06:50.120 --> 00:06:51.830 align:start position:0%
[music] ever found to fully encapsulate
all<00:06:50.400><c> of</c><00:06:50.480><c> the</c><00:06:50.560><c> dimensions</c><00:06:51.280><c> of</c><00:06:51.440><c> what</c><00:06:51.640><c> these</c>

00:06:51.830 --> 00:06:51.840 align:start position:0%
all of the dimensions of what these
 

00:06:51.840 --> 00:06:53.630 align:start position:0%
all of the dimensions of what these
companies<00:06:52.320><c> do,</c>

00:06:53.630 --> 00:06:53.640 align:start position:0%
companies do,
 

00:06:53.640 --> 00:06:56.230 align:start position:0%
companies do,
and<00:06:54.280><c> the</c><00:06:54.360><c> scale</c><00:06:54.720><c> that</c><00:06:54.840><c> they</c><00:06:54.960><c> operate,</c><00:06:55.560><c> and</c>

00:06:56.230 --> 00:06:56.240 align:start position:0%
and the scale that they operate, and
 

00:06:56.240 --> 00:06:58.390 align:start position:0%
and the scale that they operate, and
what<00:06:56.440><c> motivates</c><00:06:56.960><c> them</c><00:06:57.160><c> to</c><00:06:57.280><c> do</c><00:06:57.400><c> what</c><00:06:57.520><c> they</c><00:06:57.640><c> do.</c>

00:06:58.390 --> 00:06:58.400 align:start position:0%
what motivates them to do what they do.
 

00:06:58.400 --> 00:07:00.630 align:start position:0%
what motivates them to do what they do.
And<00:06:58.600><c> there</c><00:06:58.800><c> are</c><00:06:59.040><c> many</c><00:06:59.640><c> parallels</c><00:07:00.400><c> that</c><00:07:00.520><c> you</c>

00:07:00.630 --> 00:07:00.640 align:start position:0%
And there are many parallels that you
 

00:07:00.640 --> 00:07:02.630 align:start position:0%
And there are many parallels that you
see<00:07:01.040><c> between</c><00:07:01.680><c> what</c><00:07:01.880><c> I</c><00:07:01.919><c> call</c><00:07:02.120><c> the</c><00:07:02.200><c> empires</c><00:07:02.520><c> of</c>

00:07:02.630 --> 00:07:02.640 align:start position:0%
see between what I call the empires of
 

00:07:02.640 --> 00:07:04.870 align:start position:0%
see between what I call the empires of
AI<00:07:02.880><c> and</c><00:07:03.000><c> the</c><00:07:03.080><c> empires</c><00:07:03.480><c> of</c><00:07:03.600><c> old.</c>

00:07:04.870 --> 00:07:04.880 align:start position:0%
AI and the empires of old.
 

00:07:04.880 --> 00:07:06.670 align:start position:0%
AI and the empires of old.
&gt;&gt; Karen's<00:07:05.200><c> whole</c><00:07:05.480><c> book</c><00:07:05.760><c> rests</c><00:07:06.040><c> on</c><00:07:06.200><c> one</c><00:07:06.400><c> bold</c>

00:07:06.670 --> 00:07:06.680 align:start position:0%
&gt;&gt; Karen's whole book rests on one bold
 

00:07:06.680 --> 00:07:09.070 align:start position:0%
&gt;&gt; Karen's whole book rests on one bold
claim:<00:07:07.360><c> These</c><00:07:07.560><c> are</c><00:07:07.680><c> not</c><00:07:07.919><c> just</c><00:07:08.120><c> companies.</c><00:07:08.919><c> She</c>

00:07:09.070 --> 00:07:09.080 align:start position:0%
claim: These are not just companies. She
 

00:07:09.080 --> 00:07:11.590 align:start position:0%
claim: These are not just companies. She
calls<00:07:09.320><c> them</c><00:07:09.880><c> empires</c><00:07:10.360><c> of</c><00:07:10.520><c> AI,</c><00:07:11.160><c> and</c><00:07:11.280><c> she</c><00:07:11.400><c> lays</c>

00:07:11.590 --> 00:07:11.600 align:start position:0%
calls them empires of AI, and she lays
 

00:07:11.600 --> 00:07:13.670 align:start position:0%
calls them empires of AI, and she lays
out<00:07:11.840><c> four</c><00:07:12.120><c> specific</c><00:07:12.560><c> parallels</c><00:07:13.080><c> to</c><00:07:13.240><c> empires</c>

00:07:13.670 --> 00:07:13.680 align:start position:0%
out four specific parallels to empires
 

00:07:13.680 --> 00:07:16.710 align:start position:0%
out four specific parallels to empires
of<00:07:13.840><c> old.</c><00:07:14.760><c> One,</c><00:07:15.320><c> resource</c><00:07:15.720><c> grabbing.</c>

00:07:16.710 --> 00:07:16.720 align:start position:0%
of old. One, resource grabbing.
 

00:07:16.720 --> 00:07:18.270 align:start position:0%
of old. One, resource grabbing.
She<00:07:16.880><c> says</c><00:07:17.080><c> these</c><00:07:17.280><c> companies</c><00:07:17.680><c> lay</c><00:07:17.840><c> claim</c><00:07:18.120><c> to</c>

00:07:18.270 --> 00:07:18.280 align:start position:0%
She says these companies lay claim to
 

00:07:18.280 --> 00:07:20.110 align:start position:0%
She says these companies lay claim to
resources<00:07:18.880><c> that</c><00:07:19.080><c> aren't</c><00:07:19.320><c> theirs,</c><00:07:20.000><c> the</c>

00:07:20.110 --> 00:07:20.120 align:start position:0%
resources that aren't theirs, the
 

00:07:20.120 --> 00:07:22.350 align:start position:0%
resources that aren't theirs, the
personal<00:07:20.520><c> data</c><00:07:20.800><c> of</c><00:07:21.040><c> ordinary</c><00:07:21.480><c> people,</c><00:07:22.240><c> the</c>

00:07:22.350 --> 00:07:22.360 align:start position:0%
personal data of ordinary people, the
 

00:07:22.360 --> 00:07:23.990 align:start position:0%
personal data of ordinary people, the
intellectual<00:07:22.880><c> property</c><00:07:23.280><c> of</c><00:07:23.480><c> artists</c><00:07:23.880><c> and</c>

00:07:23.990 --> 00:07:24.000 align:start position:0%
intellectual property of artists and
 

00:07:24.000 --> 00:07:26.310 align:start position:0%
intellectual property of artists and
writers,<00:07:24.640><c> and</c><00:07:24.840><c> land</c><00:07:25.400><c> for</c><00:07:25.560><c> building</c><00:07:25.919><c> massive</c>

00:07:26.310 --> 00:07:26.320 align:start position:0%
writers, and land for building massive
 

00:07:26.320 --> 00:07:28.710 align:start position:0%
writers, and land for building massive
computing<00:07:26.760><c> facilities.</c><00:07:28.040><c> Two,</c><00:07:28.360><c> labor</c>

00:07:28.710 --> 00:07:28.720 align:start position:0%
computing facilities. Two, labor
 

00:07:28.720 --> 00:07:30.190 align:start position:0%
computing facilities. Two, labor
exploitation.

00:07:30.190 --> 00:07:30.200 align:start position:0%
exploitation.
 

00:07:30.200 --> 00:07:31.830 align:start position:0%
exploitation.
Hundreds<00:07:30.520><c> of</c><00:07:30.680><c> thousands</c><00:07:31.120><c> of</c><00:07:31.240><c> workers</c><00:07:31.600><c> around</c>

00:07:31.830 --> 00:07:31.840 align:start position:0%
Hundreds of thousands of workers around
 

00:07:31.840 --> 00:07:33.870 align:start position:0%
Hundreds of thousands of workers around
the<00:07:31.919><c> world,</c><00:07:32.320><c> including</c><00:07:32.800><c> in</c><00:07:32.880><c> the</c><00:07:33.000><c> US,</c><00:07:33.720><c> are</c>

00:07:33.870 --> 00:07:33.880 align:start position:0%
the world, including in the US, are
 

00:07:33.880 --> 00:07:36.150 align:start position:0%
the world, including in the US, are
contracted<00:07:34.560><c> to</c><00:07:34.640><c> help</c><00:07:34.960><c> build</c><00:07:35.240><c> these</c><00:07:35.440><c> systems,</c>

00:07:36.150 --> 00:07:36.160 align:start position:0%
contracted to help build these systems,
 

00:07:36.160 --> 00:07:37.909 align:start position:0%
contracted to help build these systems,
often<00:07:36.480><c> in</c><00:07:36.600><c> ways</c><00:07:36.800><c> that</c><00:07:36.960><c> quietly</c><00:07:37.440><c> strip</c><00:07:37.720><c> away</c>

00:07:37.909 --> 00:07:37.919 align:start position:0%
often in ways that quietly strip away
 

00:07:37.919 --> 00:07:39.750 align:start position:0%
often in ways that quietly strip away
labor<00:07:38.200><c> protections</c><00:07:39.080><c> once</c><00:07:39.280><c> the</c><00:07:39.400><c> tools</c><00:07:39.680><c> are</c>

00:07:39.750 --> 00:07:39.760 align:start position:0%
labor protections once the tools are
 

00:07:39.760 --> 00:07:41.150 align:start position:0%
labor protections once the tools are
deployed.

00:07:41.150 --> 00:07:41.160 align:start position:0%
deployed.
 

00:07:41.160 --> 00:07:43.230 align:start position:0%
deployed.
Three,<00:07:41.640><c> controlling</c><00:07:42.120><c> knowledge.</c>

00:07:43.230 --> 00:07:43.240 align:start position:0%
Three, controlling knowledge.
 

00:07:43.240 --> 00:07:44.910 align:start position:0%
Three, controlling knowledge.
These<00:07:43.480><c> corporate</c><00:07:43.960><c> empires</c><00:07:44.440><c> project</c><00:07:44.840><c> the</c>

00:07:44.910 --> 00:07:44.920 align:start position:0%
These corporate empires project the
 

00:07:44.920 --> 00:07:46.630 align:start position:0%
These corporate empires project the
false<00:07:45.320><c> narrative</c><00:07:45.760><c> that</c><00:07:45.920><c> they</c><00:07:46.080><c> are</c><00:07:46.200><c> the</c><00:07:46.360><c> only</c>

00:07:46.630 --> 00:07:46.640 align:start position:0%
false narrative that they are the only
 

00:07:46.640 --> 00:07:48.630 align:start position:0%
false narrative that they are the only
ones<00:07:46.920><c> pure</c><00:07:47.160><c> enough</c><00:07:47.720><c> or</c><00:07:47.880><c> smart</c><00:07:48.240><c> enough</c><00:07:48.480><c> to</c>

00:07:48.630 --> 00:07:48.640 align:start position:0%
ones pure enough or smart enough to
 

00:07:48.640 --> 00:07:51.270 align:start position:0%
ones pure enough or smart enough to
understand<00:07:49.280><c> and</c><00:07:49.440><c> guide</c><00:07:49.720><c> AI</c><00:07:50.000><c> safely.</c><00:07:51.080><c> They</c>

00:07:51.270 --> 00:07:51.280 align:start position:0%
understand and guide AI safely. They
 

00:07:51.280 --> 00:07:53.590 align:start position:0%
understand and guide AI safely. They
actively<00:07:51.800><c> gaslight</c><00:07:52.440><c> the</c><00:07:52.560><c> public</c><00:07:52.920><c> and</c><00:07:53.080><c> policy</c>

00:07:53.590 --> 00:07:53.600 align:start position:0%
actively gaslight the public and policy
 

00:07:53.600 --> 00:07:55.990 align:start position:0%
actively gaslight the public and policy
makers.<00:07:54.400><c> Even</c><00:07:54.720><c> worse,</c><00:07:55.080><c> they</c><00:07:55.240><c> spend</c><00:07:55.600><c> hundreds</c>

00:07:55.990 --> 00:07:56.000 align:start position:0%
makers. Even worse, they spend hundreds
 

00:07:56.000 --> 00:07:57.670 align:start position:0%
makers. Even worse, they spend hundreds
of<00:07:56.120><c> millions</c><00:07:56.480><c> of</c><00:07:56.600><c> dollars</c><00:07:56.960><c> to</c><00:07:57.080><c> kill</c><00:07:57.320><c> public</c>

00:07:57.670 --> 00:07:57.680 align:start position:0%
of millions of dollars to kill public
 

00:07:57.680 --> 00:07:59.950 align:start position:0%
of millions of dollars to kill public
safety<00:07:58.000><c> legislation</c><00:07:58.880><c> and</c><00:07:59.040><c> ruthlessly</c><00:07:59.560><c> censor</c>

00:07:59.950 --> 00:07:59.960 align:start position:0%
safety legislation and ruthlessly censor
 

00:07:59.960 --> 00:08:01.870 align:start position:0%
safety legislation and ruthlessly censor
any<00:08:00.160><c> internal</c><00:08:00.600><c> researchers</c><00:08:01.200><c> whose</c><00:08:01.440><c> findings</c>

00:08:01.870 --> 00:08:01.880 align:start position:0%
any internal researchers whose findings
 

00:08:01.880 --> 00:08:04.310 align:start position:0%
any internal researchers whose findings
threaten<00:08:02.160><c> their</c><00:08:02.320><c> corporate</c><00:08:02.720><c> profits.</c>

00:08:04.310 --> 00:08:04.320 align:start position:0%
threaten their corporate profits.
 

00:08:04.320 --> 00:08:06.070 align:start position:0%
threaten their corporate profits.
Take<00:08:04.520><c> the</c><00:08:04.640><c> case</c><00:08:04.880><c> of</c><00:08:05.000><c> Dr.</c><00:08:05.280><c> Timnit</c><00:08:05.640><c> Gebru</c><00:08:05.960><c> at</c>

00:08:06.070 --> 00:08:06.080 align:start position:0%
Take the case of Dr. Timnit Gebru at
 

00:08:06.080 --> 00:08:07.110 align:start position:0%
Take the case of Dr. Timnit Gebru at
Google.

00:08:07.110 --> 00:08:07.120 align:start position:0%
Google.
 

00:08:07.120 --> 00:08:09.030 align:start position:0%
Google.
She<00:08:07.240><c> was</c><00:08:07.400><c> hired</c><00:08:07.720><c> to</c><00:08:07.840><c> lead</c><00:08:08.040><c> an</c><00:08:08.200><c> ethical</c><00:08:08.680><c> AI</c>

00:08:09.030 --> 00:08:09.040 align:start position:0%
She was hired to lead an ethical AI
 

00:08:09.040 --> 00:08:10.950 align:start position:0%
She was hired to lead an ethical AI
team.<00:08:09.720><c> But</c><00:08:09.840><c> when</c><00:08:09.960><c> she</c><00:08:10.120><c> co-authored</c><00:08:10.880><c> a</c>

00:08:10.950 --> 00:08:10.960 align:start position:0%
team. But when she co-authored a
 

00:08:10.960 --> 00:08:13.270 align:start position:0%
team. But when she co-authored a
brilliant<00:08:11.480><c> paper</c><00:08:12.160><c> highlighting</c><00:08:12.680><c> the</c><00:08:12.760><c> severe</c>

00:08:13.270 --> 00:08:13.280 align:start position:0%
brilliant paper highlighting the severe
 

00:08:13.280 --> 00:08:15.630 align:start position:0%
brilliant paper highlighting the severe
biases<00:08:13.919><c> and</c><00:08:14.080><c> environmental</c><00:08:14.840><c> harms</c><00:08:15.200><c> of</c><00:08:15.320><c> large</c>

00:08:15.630 --> 00:08:15.640 align:start position:0%
biases and environmental harms of large
 

00:08:15.640 --> 00:08:18.030 align:start position:0%
biases and environmental harms of large
language<00:08:16.040><c> models,</c><00:08:16.680><c> Google</c><00:08:17.040><c> executives</c><00:08:17.720><c> tried</c>

00:08:18.030 --> 00:08:18.040 align:start position:0%
language models, Google executives tried
 

00:08:18.040 --> 00:08:20.110 align:start position:0%
language models, Google executives tried
to<00:08:18.120><c> suppress</c><00:08:18.560><c> it</c><00:08:18.840><c> and</c><00:08:19.000><c> ultimately</c><00:08:19.480><c> fired</c><00:08:19.840><c> her</c>

00:08:20.110 --> 00:08:20.120 align:start position:0%
to suppress it and ultimately fired her
 

00:08:20.120 --> 00:08:22.830 align:start position:0%
to suppress it and ultimately fired her
and<00:08:20.280><c> her</c><00:08:20.400><c> co-lead,</c><00:08:21.240><c> Margaret</c><00:08:21.720><c> Mitchell.</c>

00:08:22.830 --> 00:08:22.840 align:start position:0%
and her co-lead, Margaret Mitchell.
 

00:08:22.840 --> 00:08:25.150 align:start position:0%
and her co-lead, Margaret Mitchell.
Four,<00:08:23.400><c> myth-making.</c><00:08:24.480><c> This</c><00:08:24.640><c> is</c><00:08:24.760><c> the</c><00:08:24.840><c> one</c><00:08:25.040><c> that</c>

00:08:25.150 --> 00:08:25.160 align:start position:0%
Four, myth-making. This is the one that
 

00:08:25.160 --> 00:08:27.430 align:start position:0%
Four, myth-making. This is the one that
got<00:08:25.400><c> me.</c><00:08:25.800><c> Karen</c><00:08:26.160><c> says</c><00:08:26.520><c> these</c><00:08:26.760><c> companies</c><00:08:27.240><c> need</c>

00:08:27.430 --> 00:08:27.440 align:start position:0%
got me. Karen says these companies need
 

00:08:27.440 --> 00:08:28.950 align:start position:0%
got me. Karen says these companies need
the<00:08:27.520><c> public</c><00:08:27.880><c> to</c><00:08:27.960><c> believe</c><00:08:28.320><c> two</c><00:08:28.560><c> things</c><00:08:28.800><c> at</c>

00:08:28.950 --> 00:08:28.960 align:start position:0%
the public to believe two things at
 

00:08:28.960 --> 00:08:29.950 align:start position:0%
the public to believe two things at
once.

00:08:29.950 --> 00:08:29.960 align:start position:0%
once.
 

00:08:29.960 --> 00:08:32.190 align:start position:0%
once.
That<00:08:30.160><c> AI</c><00:08:30.520><c> could</c><00:08:30.680><c> be</c><00:08:30.800><c> catastrophic</c><00:08:31.960><c> and</c><00:08:32.080><c> that</c>

00:08:32.190 --> 00:08:32.200 align:start position:0%
That AI could be catastrophic and that
 

00:08:32.200 --> 00:08:35.110 align:start position:0%
That AI could be catastrophic and that
it<00:08:32.280><c> could</c><00:08:32.479><c> bring</c><00:08:32.880><c> utopia.</c><00:08:33.919><c> Why</c><00:08:34.120><c> both?</c><00:08:34.719><c> Because</c>

00:08:35.110 --> 00:08:35.120 align:start position:0%
it could bring utopia. Why both? Because
 

00:08:35.120 --> 00:08:37.070 align:start position:0%
it could bring utopia. Why both? Because
if<00:08:35.240><c> it's</c><00:08:35.440><c> dangerous,</c><00:08:36.320><c> only</c><00:08:36.599><c> they</c><00:08:36.800><c> should</c><00:08:36.960><c> be</c>

00:08:37.070 --> 00:08:37.080 align:start position:0%
if it's dangerous, only they should be
 

00:08:37.080 --> 00:08:38.870 align:start position:0%
if it's dangerous, only they should be
trusted<00:08:37.479><c> to</c><00:08:37.560><c> control</c><00:08:37.960><c> it.</c><00:08:38.479><c> And</c><00:08:38.599><c> if</c><00:08:38.719><c> it's</c>

00:08:38.870 --> 00:08:38.880 align:start position:0%
trusted to control it. And if it's
 

00:08:38.880 --> 00:08:40.870 align:start position:0%
trusted to control it. And if it's
utopian,<00:08:39.680><c> we</c><00:08:39.800><c> should</c><00:08:40.039><c> hand</c><00:08:40.120><c> them</c><00:08:40.240><c> our</c><00:08:40.360><c> trust,</c>

00:08:40.870 --> 00:08:40.880 align:start position:0%
utopian, we should hand them our trust,
 

00:08:40.880 --> 00:08:42.390 align:start position:0%
utopian, we should hand them our trust,
our<00:08:41.000><c> data,</c><00:08:41.479><c> and</c><00:08:41.599><c> our</c><00:08:41.719><c> money</c><00:08:42.000><c> to</c><00:08:42.080><c> get</c><00:08:42.240><c> there</c>

00:08:42.390 --> 00:08:42.400 align:start position:0%
our data, and our money to get there
 

00:08:42.400 --> 00:08:44.550 align:start position:0%
our data, and our money to get there
faster.<00:08:43.200><c> Here's</c><00:08:43.479><c> a</c><00:08:43.560><c> real</c><00:08:43.760><c> quote</c><00:08:44.080><c> she</c><00:08:44.240><c> brings</c>

00:08:44.550 --> 00:08:44.560 align:start position:0%
faster. Here's a real quote she brings
 

00:08:44.560 --> 00:08:46.310 align:start position:0%
faster. Here's a real quote she brings
up<00:08:44.839><c> from</c><00:08:45.040><c> 2017.</c>

00:08:46.310 --> 00:08:46.320 align:start position:0%
up from 2017.
 

00:08:46.320 --> 00:08:48.150 align:start position:0%
up from 2017.
&gt;&gt; the<00:08:46.400><c> record</c><00:08:46.760><c> on</c><00:08:46.880><c> podcasts</c><00:08:47.680><c> in</c><00:08:47.800><c> their</c><00:08:47.920><c> blog</c>

00:08:48.150 --> 00:08:48.160 align:start position:0%
&gt;&gt; the record on podcasts in their blog
 

00:08:48.160 --> 00:08:49.990 align:start position:0%
&gt;&gt; the record on podcasts in their blog
posts<00:08:48.480><c> to</c><00:08:48.600><c> see</c><00:08:48.800><c> how</c><00:08:49.240><c> it's</c><00:08:49.400><c> evolved</c><00:08:49.680><c> over</c><00:08:49.839><c> time.</c>

00:08:49.990 --> 00:08:50.000 align:start position:0%
posts to see how it's evolved over time.
 

00:08:50.000 --> 00:08:51.790 align:start position:0%
posts to see how it's evolved over time.
And<00:08:50.080><c> Dario,</c><00:08:50.560><c> who</c><00:08:50.720><c> was</c><00:08:50.880><c> the</c><00:08:50.960><c> former</c><00:08:51.320><c> VP</c><00:08:51.680><c> of</c>

00:08:51.790 --> 00:08:51.800 align:start position:0%
And Dario, who was the former VP of
 

00:08:51.800 --> 00:08:54.230 align:start position:0%
And Dario, who was the former VP of
research<00:08:52.360><c> at</c><00:08:52.480><c> OpenAI,</c><00:08:53.440><c> and</c><00:08:53.600><c> has</c><00:08:53.720><c> now</c><00:08:53.920><c> moved</c><00:08:54.120><c> on</c>

00:08:54.230 --> 00:08:54.240 align:start position:0%
research at OpenAI, and has now moved on
 

00:08:54.240 --> 00:08:56.070 align:start position:0%
research at OpenAI, and has now moved on
to<00:08:54.360><c> Anthropic,</c><00:08:55.240><c> who</c><00:08:55.360><c> are</c><00:08:55.400><c> taking</c><00:08:55.680><c> a</c><00:08:55.720><c> slightly</c>

00:08:56.070 --> 00:08:56.080 align:start position:0%
to Anthropic, who are taking a slightly
 

00:08:56.080 --> 00:08:58.070 align:start position:0%
to Anthropic, who are taking a slightly
different<00:08:56.440><c> approach</c><00:08:56.760><c> to</c><00:08:56.880><c> developing</c><00:08:57.360><c> AI,</c>

00:08:58.070 --> 00:08:58.080 align:start position:0%
different approach to developing AI,
 

00:08:58.080 --> 00:09:00.350 align:start position:0%
different approach to developing AI,
said<00:08:58.520><c> back</c><00:08:58.800><c> in</c><00:08:58.920><c> 2017</c><00:08:59.800><c> while</c><00:08:59.920><c> he</c><00:09:00.000><c> was</c><00:09:00.120><c> still</c><00:09:00.280><c> at</c>

00:09:00.350 --> 00:09:00.360 align:start position:0%
said back in 2017 while he was still at
 

00:09:00.360 --> 00:09:03.829 align:start position:0%
said back in 2017 while he was still at
OpenAI,<00:09:01.520><c> that</c><00:09:02.240><c> I</c><00:09:02.360><c> think</c><00:09:02.560><c> at</c><00:09:02.640><c> the</c><00:09:02.720><c> extreme</c><00:09:03.320><c> end</c>

00:09:03.829 --> 00:09:03.839 align:start position:0%
OpenAI, that I think at the extreme end
 

00:09:03.839 --> 00:09:05.870 align:start position:0%
OpenAI, that I think at the extreme end
is<00:09:04.000><c> the</c><00:09:04.120><c> Nick</c><00:09:04.400><c> Bostrom</c><00:09:04.839><c> style</c><00:09:05.120><c> of</c><00:09:05.240><c> fear</c><00:09:05.760><c> that</c>

00:09:05.870 --> 00:09:05.880 align:start position:0%
is the Nick Bostrom style of fear that
 

00:09:05.880 --> 00:09:08.710 align:start position:0%
is the Nick Bostrom style of fear that
an<00:09:06.000><c> AGI</c><00:09:06.440><c> could</c><00:09:06.600><c> destroy</c><00:09:06.880><c> humanity.</c><00:09:07.920><c> I</c><00:09:08.040><c> can't</c>

00:09:08.710 --> 00:09:08.720 align:start position:0%
an AGI could destroy humanity. I can't
 

00:09:08.720 --> 00:09:11.070 align:start position:0%
an AGI could destroy humanity. I can't
see<00:09:09.040><c> any</c><00:09:09.320><c> reason</c><00:09:09.680><c> in</c><00:09:09.800><c> principle</c><00:09:10.760><c> why</c><00:09:10.880><c> that</c>

00:09:11.070 --> 00:09:11.080 align:start position:0%
see any reason in principle why that
 

00:09:11.080 --> 00:09:12.990 align:start position:0%
see any reason in principle why that
couldn't<00:09:11.320><c> happen.</c><00:09:12.080><c> My</c><00:09:12.320><c> chance</c><00:09:12.839><c> that</c>

00:09:12.990 --> 00:09:13.000 align:start position:0%
couldn't happen. My chance that
 

00:09:13.000 --> 00:09:14.510 align:start position:0%
couldn't happen. My chance that
something<00:09:13.400><c> goes</c><00:09:13.880><c> really</c><00:09:14.160><c> quite</c>

00:09:14.510 --> 00:09:14.520 align:start position:0%
something goes really quite
 

00:09:14.520 --> 00:09:16.150 align:start position:0%
something goes really quite
catastrophically<00:09:15.320><c> wrong</c><00:09:15.600><c> on</c><00:09:15.720><c> the</c><00:09:15.800><c> scale</c><00:09:16.040><c> of</c>

00:09:16.150 --> 00:09:16.160 align:start position:0%
catastrophically wrong on the scale of
 

00:09:16.160 --> 00:09:17.950 align:start position:0%
catastrophically wrong on the scale of
human<00:09:16.400><c> civilization</c>

00:09:17.950 --> 00:09:17.960 align:start position:0%
human civilization
 

00:09:17.960 --> 00:09:21.670 align:start position:0%
human civilization
might<00:09:18.280><c> be</c><00:09:18.400><c> somewhere</c><00:09:18.720><c> between</c><00:09:19.120><c> 10%</c><00:09:19.839><c> and</c><00:09:19.920><c> 25%.</c>

00:09:21.670 --> 00:09:21.680 align:start position:0%
might be somewhere between 10% and 25%.
 

00:09:21.680 --> 00:09:24.790 align:start position:0%
might be somewhere between 10% and 25%.
&gt;&gt; 10<00:09:21.920><c> to</c><00:09:22.040><c> 25%.</c><00:09:23.480><c> Read</c><00:09:23.680><c> that</c><00:09:23.839><c> number</c><00:09:24.120><c> again.</c>

00:09:24.790 --> 00:09:24.800 align:start position:0%
&gt;&gt; 10 to 25%. Read that number again.
 

00:09:24.800 --> 00:09:26.750 align:start position:0%
&gt;&gt; 10 to 25%. Read that number again.
That's<00:09:25.080><c> not</c><00:09:25.320><c> a</c><00:09:25.400><c> fringe</c><00:09:25.720><c> blogger.</c><00:09:26.520><c> That's</c><00:09:26.680><c> a</c>

00:09:26.750 --> 00:09:26.760 align:start position:0%
That's not a fringe blogger. That's a
 

00:09:26.760 --> 00:09:28.670 align:start position:0%
That's not a fringe blogger. That's a
man<00:09:26.960><c> now</c><00:09:27.200><c> running</c><00:09:27.520><c> one</c><00:09:27.640><c> of</c><00:09:27.720><c> the</c><00:09:27.800><c> most</c><00:09:28.080><c> funded</c>

00:09:28.670 --> 00:09:28.680 align:start position:0%
man now running one of the most funded
 

00:09:28.680 --> 00:09:31.070 align:start position:0%
man now running one of the most funded
AI<00:09:29.040><c> labs</c><00:09:29.440><c> on</c><00:09:29.600><c> Earth.</c>

00:09:31.070 --> 00:09:31.080 align:start position:0%
AI labs on Earth.
 

00:09:31.080 --> 00:09:32.670 align:start position:0%
AI labs on Earth.
I<00:09:31.120><c> thought</c><00:09:31.360><c> the</c><00:09:31.480><c> empire</c><00:09:31.920><c> argument</c><00:09:32.440><c> was</c><00:09:32.600><c> the</c>

00:09:32.670 --> 00:09:32.680 align:start position:0%
I thought the empire argument was the
 

00:09:32.680 --> 00:09:35.350 align:start position:0%
I thought the empire argument was the
most<00:09:32.920><c> shocking</c><00:09:33.360><c> part</c><00:09:33.720><c> of</c><00:09:33.800><c> this</c><00:09:33.920><c> conversation.</c>

00:09:35.350 --> 00:09:35.360 align:start position:0%
most shocking part of this conversation.
 

00:09:35.360 --> 00:09:37.670 align:start position:0%
most shocking part of this conversation.
I<00:09:35.440><c> was</c><00:09:35.600><c> wrong.</c><00:09:36.480><c> The</c><00:09:36.520><c> next</c><00:09:36.760><c> section</c><00:09:37.360><c> about</c><00:09:37.600><c> a</c>

00:09:37.670 --> 00:09:37.680 align:start position:0%
I was wrong. The next section about a
 

00:09:37.680 --> 00:09:39.990 align:start position:0%
I was wrong. The next section about a
small<00:09:37.960><c> community</c><00:09:38.360><c> in</c><00:09:38.440><c> Tennessee</c><00:09:39.400><c> is</c><00:09:39.560><c> the</c><00:09:39.680><c> part</c>

00:09:39.990 --> 00:09:40.000 align:start position:0%
small community in Tennessee is the part
 

00:09:40.000 --> 00:09:42.110 align:start position:0%
small community in Tennessee is the part
that<00:09:40.160><c> actually</c><00:09:40.480><c> stayed</c><00:09:40.840><c> with</c><00:09:40.960><c> me</c><00:09:41.120><c> overnight.</c>

00:09:42.110 --> 00:09:42.120 align:start position:0%
that actually stayed with me overnight.
 

00:09:42.120 --> 00:09:44.470 align:start position:0%
that actually stayed with me overnight.
We<00:09:42.200><c> talk</c><00:09:42.480><c> about</c><00:09:42.840><c> AI</c><00:09:43.320><c> like</c><00:09:43.520><c> it</c><00:09:43.640><c> lives</c><00:09:43.920><c> in</c><00:09:44.360><c> the</c>

00:09:44.470 --> 00:09:44.480 align:start position:0%
We talk about AI like it lives in the
 

00:09:44.480 --> 00:09:47.710 align:start position:0%
We talk about AI like it lives in the
cloud,<00:09:45.280><c> like</c><00:09:45.440><c> it's</c><00:09:45.640><c> weightless.</c><00:09:46.560><c> It</c><00:09:46.720><c> is</c><00:09:46.840><c> not.</c>

00:09:47.710 --> 00:09:47.720 align:start position:0%
cloud, like it's weightless. It is not.
 

00:09:47.720 --> 00:09:51.030 align:start position:0%
cloud, like it's weightless. It is not.
&gt;&gt; OpenAI's<00:09:49.120><c> largest,</c><00:09:50.000><c> one</c><00:09:50.160><c> of</c><00:09:50.280><c> its</c><00:09:50.480><c> largest</c>

00:09:51.030 --> 00:09:51.040 align:start position:0%
&gt;&gt; OpenAI's largest, one of its largest
 

00:09:51.040 --> 00:09:52.870 align:start position:0%
&gt;&gt; OpenAI's largest, one of its largest
data<00:09:51.200><c> center</c><00:09:51.440><c> projects,</c><00:09:52.120><c> is</c><00:09:52.240><c> being</c><00:09:52.480><c> built</c><00:09:52.720><c> in</c>

00:09:52.870 --> 00:09:52.880 align:start position:0%
data center projects, is being built in
 

00:09:52.880 --> 00:09:54.630 align:start position:0%
data center projects, is being built in
Abilene,<00:09:53.240><c> Texas</c><00:09:53.680><c> as</c><00:09:53.800><c> part</c><00:09:54.000><c> of</c><00:09:54.080><c> the</c><00:09:54.160><c> Stargate</c>

00:09:54.630 --> 00:09:54.640 align:start position:0%
Abilene, Texas as part of the Stargate
 

00:09:54.640 --> 00:09:56.790 align:start position:0%
Abilene, Texas as part of the Stargate
initiative,<00:09:55.160><c> which</c><00:09:55.400><c> was</c><00:09:56.280><c> an</c><00:09:56.480><c> effort</c>

00:09:56.790 --> 00:09:56.800 align:start position:0%
initiative, which was an effort
 

00:09:56.800 --> 00:09:58.590 align:start position:0%
initiative, which was an effort
announced<00:09:57.160><c> at</c><00:09:57.240><c> the</c><00:09:57.320><c> beginning</c><00:09:57.640><c> of</c><00:09:58.320><c> Trump's</c>

00:09:58.590 --> 00:09:58.600 align:start position:0%
announced at the beginning of Trump's
 

00:09:58.600 --> 00:10:00.310 align:start position:0%
announced at the beginning of Trump's
second<00:09:58.840><c> administration</c><00:09:59.440><c> to</c><00:09:59.560><c> spend</c><00:09:59.880><c> $500</c>

00:10:00.310 --> 00:10:00.320 align:start position:0%
second administration to spend $500
 

00:10:00.320 --> 00:10:03.830 align:start position:0%
second administration to spend $500
billion<00:10:01.040><c> on</c><00:10:01.400><c> AI</c><00:10:01.680><c> computing</c><00:10:02.080><c> infrastructure.</c>

00:10:03.830 --> 00:10:03.840 align:start position:0%
billion on AI computing infrastructure.
 

00:10:03.840 --> 00:10:05.630 align:start position:0%
billion on AI computing infrastructure.
That<00:10:04.120><c> one</c><00:10:04.560><c> would</c><00:10:04.720><c> be</c><00:10:04.800><c> the</c><00:10:04.880><c> size</c><00:10:05.280><c> of</c><00:10:05.360><c> Central</c>

00:10:05.630 --> 00:10:05.640 align:start position:0%
That one would be the size of Central
 

00:10:05.640 --> 00:10:06.910 align:start position:0%
That one would be the size of Central
Park.

00:10:06.910 --> 00:10:06.920 align:start position:0%
Park.
 

00:10:06.920 --> 00:10:09.070 align:start position:0%
Park.
And<00:10:07.360><c> it</c><00:10:07.480><c> would</c><00:10:07.680><c> run</c><00:10:08.080><c> a</c><00:10:08.160><c> million</c><00:10:08.680><c> computer</c>

00:10:09.070 --> 00:10:09.080 align:start position:0%
And it would run a million computer
 

00:10:09.080 --> 00:10:11.750 align:start position:0%
And it would run a million computer
chips,<00:10:10.000><c> and</c><00:10:10.320><c> it</c><00:10:10.480><c> would</c>

00:10:11.750 --> 00:10:11.760 align:start position:0%
chips, and it would
 

00:10:11.760 --> 00:10:14.750 align:start position:0%
chips, and it would
require<00:10:12.480><c> the</c><00:10:12.600><c> power</c><00:10:13.240><c> of</c><00:10:13.520><c> more</c><00:10:13.680><c> than</c><00:10:13.800><c> 20%</c><00:10:14.640><c> of</c>

00:10:14.750 --> 00:10:14.760 align:start position:0%
require the power of more than 20% of
 

00:10:14.760 --> 00:10:16.350 align:start position:0%
require the power of more than 20% of
New<00:10:14.840><c> York</c><00:10:15.040><c> City.</c>

00:10:16.350 --> 00:10:16.360 align:start position:0%
New York City.
 

00:10:16.360 --> 00:10:18.110 align:start position:0%
New York City.
&gt;&gt; Then<00:10:16.480><c> there's</c><00:10:16.680><c> a</c><00:10:16.720><c> meta</c><00:10:17.000><c> facility</c><00:10:17.560><c> being</c><00:10:17.880><c> built</c>

00:10:18.110 --> 00:10:18.120 align:start position:0%
&gt;&gt; Then there's a meta facility being built
 

00:10:18.120 --> 00:10:20.390 align:start position:0%
&gt;&gt; Then there's a meta facility being built
in<00:10:18.240><c> Louisiana</c><00:10:19.040><c> that's</c><00:10:19.280><c> reportedly</c><00:10:20.040><c> four</c>

00:10:20.390 --> 00:10:20.400 align:start position:0%
in Louisiana that's reportedly four
 

00:10:20.400 --> 00:10:22.630 align:start position:0%
in Louisiana that's reportedly four
times<00:10:20.720><c> the</c><00:10:20.840><c> size</c><00:10:21.160><c> of</c><00:10:21.240><c> that</c><00:10:21.480><c> Texas</c><00:10:21.880><c> one,</c><00:10:22.400><c> using</c>

00:10:22.630 --> 00:10:22.640 align:start position:0%
times the size of that Texas one, using
 

00:10:22.640 --> 00:10:24.510 align:start position:0%
times the size of that Texas one, using
about<00:10:22.960><c> half</c><00:10:23.240><c> the</c><00:10:23.400><c> average</c><00:10:23.720><c> power</c><00:10:24.040><c> demand</c><00:10:24.440><c> of</c>

00:10:24.510 --> 00:10:24.520 align:start position:0%
about half the average power demand of
 

00:10:24.520 --> 00:10:26.910 align:start position:0%
about half the average power demand of
the<00:10:24.600><c> entire</c><00:10:25.080><c> city</c><00:10:25.440><c> of</c><00:10:25.560><c> New</c><00:10:25.720><c> York.</c>

00:10:26.910 --> 00:10:26.920 align:start position:0%
the entire city of New York.
 

00:10:26.920 --> 00:10:28.870 align:start position:0%
the entire city of New York.
Karen<00:10:27.200><c> talks</c><00:10:27.440><c> about</c><00:10:27.680><c> Memphis,</c><00:10:28.040><c> Tennessee,</c>

00:10:28.870 --> 00:10:28.880 align:start position:0%
Karen talks about Memphis, Tennessee,
 

00:10:28.880 --> 00:10:30.990 align:start position:0%
Karen talks about Memphis, Tennessee,
where<00:10:29.080><c> Elon</c><00:10:29.400><c> Musk</c><00:10:29.760><c> built</c><00:10:30.000><c> Colossus,</c><00:10:30.920><c> the</c>

00:10:30.990 --> 00:10:31.000 align:start position:0%
where Elon Musk built Colossus, the
 

00:10:31.000 --> 00:10:33.030 align:start position:0%
where Elon Musk built Colossus, the
supercomputer<00:10:31.800><c> used</c><00:10:32.040><c> to</c><00:10:32.160><c> train</c><00:10:32.440><c> Grok,</c>

00:10:33.030 --> 00:10:33.040 align:start position:0%
supercomputer used to train Grok,
 

00:10:33.040 --> 00:10:35.310 align:start position:0%
supercomputer used to train Grok,
powered<00:10:33.360><c> partly</c><00:10:33.720><c> by</c><00:10:33.880><c> 35</c><00:10:34.520><c> methane</c><00:10:35.080><c> gas</c>

00:10:35.310 --> 00:10:35.320 align:start position:0%
powered partly by 35 methane gas
 

00:10:35.320 --> 00:10:36.390 align:start position:0%
powered partly by 35 methane gas
turbines.

00:10:36.390 --> 00:10:36.400 align:start position:0%
turbines.
 

00:10:36.400 --> 00:10:38.190 align:start position:0%
turbines.
The<00:10:36.440><c> community</c><00:10:37.040><c> living</c><00:10:37.440><c> there,</c><00:10:37.720><c> a</c><00:10:37.800><c> working</c>

00:10:38.190 --> 00:10:38.200 align:start position:0%
The community living there, a working
 

00:10:38.200 --> 00:10:40.790 align:start position:0%
The community living there, a working
class,<00:10:38.920><c> reportedly</c><00:10:39.480><c> wasn't</c><00:10:39.880><c> formally</c><00:10:40.360><c> told</c>

00:10:40.790 --> 00:10:40.800 align:start position:0%
class, reportedly wasn't formally told
 

00:10:40.800 --> 00:10:43.110 align:start position:0%
class, reportedly wasn't formally told
they'd<00:10:40.960><c> be</c><00:10:41.120><c> hosting</c><00:10:41.640><c> this</c><00:10:41.839><c> facility.</c>

00:10:43.110 --> 00:10:43.120 align:start position:0%
they'd be hosting this facility.
 

00:10:43.120 --> 00:10:44.870 align:start position:0%
they'd be hosting this facility.
According<00:10:43.400><c> to</c><00:10:43.520><c> Karen,</c><00:10:44.080><c> residents</c><00:10:44.560><c> first</c>

00:10:44.870 --> 00:10:44.880 align:start position:0%
According to Karen, residents first
 

00:10:44.880 --> 00:10:46.550 align:start position:0%
According to Karen, residents first
realized<00:10:45.240><c> something</c><00:10:45.560><c> had</c><00:10:45.720><c> changed</c><00:10:46.360><c> when</c><00:10:46.480><c> they</c>

00:10:46.550 --> 00:10:46.560 align:start position:0%
realized something had changed when they
 

00:10:46.560 --> 00:10:48.190 align:start position:0%
realized something had changed when they
smelled<00:10:46.880><c> what</c><00:10:47.040><c> seemed</c><00:10:47.240><c> like</c><00:10:47.440><c> a</c><00:10:47.520><c> gas</c><00:10:47.960><c> leak</c>

00:10:48.190 --> 00:10:48.200 align:start position:0%
smelled what seemed like a gas leak
 

00:10:48.200 --> 00:10:49.910 align:start position:0%
smelled what seemed like a gas leak
inside<00:10:48.600><c> their</c><00:10:48.760><c> own</c><00:10:48.920><c> homes.</c>

00:10:49.910 --> 00:10:49.920 align:start position:0%
inside their own homes.
 

00:10:49.920 --> 00:10:51.910 align:start position:0%
inside their own homes.
That<00:10:50.120><c> community,</c><00:10:50.600><c> she</c><00:10:50.760><c> says,</c><00:10:51.240><c> already</c><00:10:51.720><c> has</c>

00:10:51.910 --> 00:10:51.920 align:start position:0%
That community, she says, already has
 

00:10:51.920 --> 00:10:53.750 align:start position:0%
That community, she says, already has
some<00:10:52.080><c> of</c><00:10:52.160><c> the</c><00:10:52.280><c> highest</c><00:10:52.760><c> rates</c><00:10:52.960><c> of</c><00:10:53.160><c> asthma</c><00:10:53.600><c> and</c>

00:10:53.750 --> 00:10:53.760 align:start position:0%
some of the highest rates of asthma and
 

00:10:53.760 --> 00:10:56.150 align:start position:0%
some of the highest rates of asthma and
respiratory<00:10:54.400><c> illness</c><00:10:54.760><c> in</c><00:10:54.839><c> the</c><00:10:54.920><c> area.</c>

00:10:56.150 --> 00:10:56.160 align:start position:0%
respiratory illness in the area.
 

00:10:56.160 --> 00:10:58.430 align:start position:0%
respiratory illness in the area.
Now<00:10:56.400><c> add</c><00:10:56.560><c> industrial-scale</c><00:10:57.520><c> gas</c><00:10:57.839><c> turbines</c>

00:10:58.430 --> 00:10:58.440 align:start position:0%
Now add industrial-scale gas turbines
 

00:10:58.440 --> 00:11:01.150 align:start position:0%
Now add industrial-scale gas turbines
running<00:10:58.720><c> next</c><00:10:59.000><c> door,</c><00:10:59.600><c> 24</c><00:11:00.160><c> hours</c><00:11:00.400><c> a</c><00:11:00.480><c> day,</c><00:11:01.040><c> to</c>

00:11:01.150 --> 00:11:01.160 align:start position:0%
running next door, 24 hours a day, to
 

00:11:01.160 --> 00:11:03.070 align:start position:0%
running next door, 24 hours a day, to
power<00:11:01.480><c> a</c><00:11:01.560><c> machine</c><00:11:02.120><c> most</c><00:11:02.400><c> of</c><00:11:02.440><c> those</c><00:11:02.640><c> residents</c>

00:11:03.070 --> 00:11:03.080 align:start position:0%
power a machine most of those residents
 

00:11:03.080 --> 00:11:05.510 align:start position:0%
power a machine most of those residents
will<00:11:03.240><c> never</c><00:11:03.600><c> directly</c><00:11:04.040><c> benefit</c><00:11:04.480><c> from.</c>

00:11:05.510 --> 00:11:05.520 align:start position:0%
will never directly benefit from.
 

00:11:05.520 --> 00:11:07.910 align:start position:0%
will never directly benefit from.
And<00:11:05.640><c> in</c><00:11:05.760><c> a</c><00:11:05.800><c> cruel</c><00:11:06.160><c> twist,</c><00:11:06.680><c> the</c><00:11:06.839><c> same</c><00:11:07.360><c> wave</c><00:11:07.760><c> of</c>

00:11:07.910 --> 00:11:07.920 align:start position:0%
And in a cruel twist, the same wave of
 

00:11:07.920 --> 00:11:10.310 align:start position:0%
And in a cruel twist, the same wave of
automation<00:11:08.520><c> this</c><00:11:08.720><c> facility</c><00:11:09.200><c> helps</c><00:11:09.480><c> build</c><00:11:10.120><c> is</c>

00:11:10.310 --> 00:11:10.320 align:start position:0%
automation this facility helps build is
 

00:11:10.320 --> 00:11:12.150 align:start position:0%
automation this facility helps build is
also<00:11:10.640><c> the</c><00:11:10.760><c> thing</c><00:11:11.040><c> most</c><00:11:11.320><c> likely</c><00:11:11.640><c> to</c><00:11:11.760><c> replace</c>

00:11:12.150 --> 00:11:12.160 align:start position:0%
also the thing most likely to replace
 

00:11:12.160 --> 00:11:14.790 align:start position:0%
also the thing most likely to replace
the<00:11:12.280><c> jobs</c><00:11:12.720><c> many</c><00:11:12.960><c> of</c><00:11:13.040><c> them</c><00:11:13.280><c> do.</c>

00:11:14.790 --> 00:11:14.800 align:start position:0%
the jobs many of them do.
 

00:11:14.800 --> 00:11:16.630 align:start position:0%
the jobs many of them do.
I<00:11:14.880><c> want</c><00:11:15.040><c> to</c><00:11:15.080><c> be</c><00:11:15.200><c> fair</c><00:11:15.480><c> here,</c><00:11:16.080><c> because</c><00:11:16.320><c> I</c><00:11:16.400><c> think</c>

00:11:16.630 --> 00:11:16.640 align:start position:0%
I want to be fair here, because I think
 

00:11:16.640 --> 00:11:19.270 align:start position:0%
I want to be fair here, because I think
it<00:11:16.720><c> matters.</c><00:11:17.560><c> AI</c><00:11:17.880><c> data</c><00:11:18.160><c> centers</c><00:11:18.560><c> also</c><00:11:18.880><c> create</c>

00:11:19.270 --> 00:11:19.280 align:start position:0%
it matters. AI data centers also create
 

00:11:19.280 --> 00:11:22.150 align:start position:0%
it matters. AI data centers also create
construction<00:11:19.880><c> jobs,</c><00:11:20.520><c> local</c><00:11:20.880><c> tax</c><00:11:21.320><c> revenue,</c>

00:11:22.150 --> 00:11:22.160 align:start position:0%
construction jobs, local tax revenue,
 

00:11:22.160 --> 00:11:24.350 align:start position:0%
construction jobs, local tax revenue,
and<00:11:22.280><c> in</c><00:11:22.400><c> some</c><00:11:22.640><c> cases</c><00:11:23.360><c> genuinely</c><00:11:23.960><c> useful</c>

00:11:24.350 --> 00:11:24.360 align:start position:0%
and in some cases genuinely useful
 

00:11:24.360 --> 00:11:26.630 align:start position:0%
and in some cases genuinely useful
technology<00:11:24.920><c> for</c><00:11:25.080><c> medicine</c><00:11:25.480><c> and</c><00:11:25.640><c> science.</c>

00:11:26.630 --> 00:11:26.640 align:start position:0%
technology for medicine and science.
 

00:11:26.640 --> 00:11:29.150 align:start position:0%
technology for medicine and science.
This<00:11:26.800><c> isn't</c><00:11:27.040><c> a</c><00:11:27.120><c> simple</c><00:11:27.480><c> villain</c><00:11:27.839><c> story.</c>

00:11:29.150 --> 00:11:29.160 align:start position:0%
This isn't a simple villain story.
 

00:11:29.160 --> 00:11:30.829 align:start position:0%
This isn't a simple villain story.
Karen<00:11:29.440><c> says</c><00:11:29.680><c> these</c><00:11:29.920><c> companies</c><00:11:30.400><c> pick</c><00:11:30.600><c> some</c><00:11:30.760><c> of</c>

00:11:30.829 --> 00:11:30.839 align:start position:0%
Karen says these companies pick some of
 

00:11:30.839 --> 00:11:32.390 align:start position:0%
Karen says these companies pick some of
the<00:11:30.920><c> most</c><00:11:31.160><c> vulnerable</c><00:11:31.680><c> communities</c><00:11:32.280><c> for</c>

00:11:32.390 --> 00:11:32.400 align:start position:0%
the most vulnerable communities for
 

00:11:32.400 --> 00:11:34.230 align:start position:0%
the most vulnerable communities for
these<00:11:32.600><c> projects.</c><00:11:33.480><c> And</c><00:11:33.600><c> when</c><00:11:33.760><c> you</c><00:11:33.880><c> look</c><00:11:34.080><c> at</c>

00:11:34.230 --> 00:11:34.240 align:start position:0%
these projects. And when you look at
 

00:11:34.240 --> 00:11:35.870 align:start position:0%
these projects. And when you look at
where<00:11:34.400><c> facilities</c><00:11:35.000><c> like</c><00:11:35.200><c> this</c><00:11:35.480><c> actually</c>

00:11:35.870 --> 00:11:35.880 align:start position:0%
where facilities like this actually
 

00:11:35.880 --> 00:11:37.630 align:start position:0%
where facilities like this actually
land,<00:11:36.640><c> it's</c><00:11:36.840><c> hard</c><00:11:37.120><c> to</c><00:11:37.200><c> call</c><00:11:37.440><c> that</c><00:11:37.600><c> a</c>

00:11:37.630 --> 00:11:37.640 align:start position:0%
land, it's hard to call that a
 

00:11:37.640 --> 00:11:39.670 align:start position:0%
land, it's hard to call that a
coincidence,<00:11:38.360><c> either.</c>

00:11:39.670 --> 00:11:39.680 align:start position:0%
coincidence, either.
 

00:11:39.680 --> 00:11:40.910 align:start position:0%
coincidence, either.
Now,<00:11:39.800><c> here's</c><00:11:40.080><c> the</c><00:11:40.200><c> part</c><00:11:40.480><c> where</c><00:11:40.600><c> you</c><00:11:40.720><c> might</c>

00:11:40.910 --> 00:11:40.920 align:start position:0%
Now, here's the part where you might
 

00:11:40.920 --> 00:11:42.710 align:start position:0%
Now, here's the part where you might
expect<00:11:41.320><c> me</c><00:11:41.440><c> to</c><00:11:41.520><c> just</c><00:11:41.760><c> tell</c><00:11:41.960><c> you</c><00:11:42.040><c> to</c><00:11:42.160><c> log</c><00:11:42.480><c> off</c>

00:11:42.710 --> 00:11:42.720 align:start position:0%
expect me to just tell you to log off
 

00:11:42.720 --> 00:11:44.910 align:start position:0%
expect me to just tell you to log off
and<00:11:42.840><c> boycott</c><00:11:43.400><c> AI</c><00:11:43.720><c> forever.</c>

00:11:44.910 --> 00:11:44.920 align:start position:0%
and boycott AI forever.
 

00:11:44.920 --> 00:11:46.670 align:start position:0%
and boycott AI forever.
That's<00:11:45.200><c> not</c><00:11:45.400><c> what</c><00:11:45.560><c> Karen</c><00:11:45.840><c> says.</c><00:11:46.320><c> Her</c><00:11:46.440><c> answer</c>

00:11:46.670 --> 00:11:46.680 align:start position:0%
That's not what Karen says. Her answer
 

00:11:46.680 --> 00:11:48.750 align:start position:0%
That's not what Karen says. Her answer
is<00:11:46.800><c> more</c><00:11:47.040><c> interesting</c><00:11:47.760><c> and</c><00:11:47.960><c> honestly</c><00:11:48.440><c> more</c>

00:11:48.750 --> 00:11:48.760 align:start position:0%
is more interesting and honestly more
 

00:11:48.760 --> 00:11:50.030 align:start position:0%
is more interesting and honestly more
useful.

00:11:50.030 --> 00:11:50.040 align:start position:0%
useful.
 

00:11:50.040 --> 00:11:52.150 align:start position:0%
useful.
Karen's<00:11:50.360><c> suggestion</c><00:11:50.880><c> isn't</c><00:11:51.200><c> delete</c><00:11:51.560><c> every</c><00:11:51.840><c> AI</c>

00:11:52.150 --> 00:11:52.160 align:start position:0%
Karen's suggestion isn't delete every AI
 

00:11:52.160 --> 00:11:54.790 align:start position:0%
Karen's suggestion isn't delete every AI
app.<00:11:52.720><c> It's</c><00:11:52.960><c> smarter</c><00:11:53.360><c> than</c><00:11:53.520><c> that.</c><00:11:54.320><c> She</c><00:11:54.440><c> calls</c>

00:11:54.790 --> 00:11:54.800 align:start position:0%
app. It's smarter than that. She calls
 

00:11:54.800 --> 00:11:57.150 align:start position:0%
app. It's smarter than that. She calls
all<00:11:55.080><c> of</c><00:11:55.200><c> us</c><00:11:55.400><c> data</c><00:11:55.680><c> donors,</c><00:11:56.480><c> meaning</c><00:11:56.880><c> every</c>

00:11:57.150 --> 00:11:57.160 align:start position:0%
all of us data donors, meaning every
 

00:11:57.160 --> 00:11:59.510 align:start position:0%
all of us data donors, meaning every
time<00:11:57.400><c> we</c><00:11:57.560><c> use</c><00:11:57.840><c> these</c><00:11:58.080><c> tools,</c><00:11:59.040><c> we're</c><00:11:59.160><c> feeding</c>

00:11:59.510 --> 00:11:59.520 align:start position:0%
time we use these tools, we're feeding
 

00:11:59.520 --> 00:12:01.350 align:start position:0%
time we use these tools, we're feeding
the<00:11:59.600><c> very</c><00:11:59.880><c> systems</c><00:12:00.360><c> these</c><00:12:00.560><c> companies</c><00:12:01.040><c> need</c><00:12:01.280><c> to</c>

00:12:01.350 --> 00:12:01.360 align:start position:0%
the very systems these companies need to
 

00:12:01.360 --> 00:12:03.790 align:start position:0%
the very systems these companies need to
keep<00:12:01.640><c> growing.</c><00:12:02.480><c> Her</c><00:12:02.640><c> point,</c><00:12:03.280><c> you</c><00:12:03.400><c> have</c><00:12:03.560><c> more</c>

00:12:03.790 --> 00:12:03.800 align:start position:0%
keep growing. Her point, you have more
 

00:12:03.800 --> 00:12:05.870 align:start position:0%
keep growing. Her point, you have more
leverage<00:12:04.240><c> than</c><00:12:04.400><c> you</c><00:12:04.480><c> think.</c><00:12:05.240><c> She</c><00:12:05.400><c> lists</c><00:12:05.640><c> a</c><00:12:05.680><c> few</c>

00:12:05.870 --> 00:12:05.880 align:start position:0%
leverage than you think. She lists a few
 

00:12:05.880 --> 00:12:07.510 align:start position:0%
leverage than you think. She lists a few
concrete<00:12:06.360><c> things,</c><00:12:06.960><c> and</c><00:12:07.080><c> I</c><00:12:07.120><c> want</c><00:12:07.240><c> to</c><00:12:07.320><c> walk</c>

00:12:07.510 --> 00:12:07.520 align:start position:0%
concrete things, and I want to walk
 

00:12:07.520 --> 00:12:09.430 align:start position:0%
concrete things, and I want to walk
through<00:12:07.640><c> them</c><00:12:07.800><c> simply.</c><00:12:08.720><c> Withhold</c><00:12:09.200><c> what</c><00:12:09.320><c> you</c>

00:12:09.430 --> 00:12:09.440 align:start position:0%
through them simply. Withhold what you
 

00:12:09.440 --> 00:12:11.630 align:start position:0%
through them simply. Withhold what you
can.<00:12:10.200><c> Artists</c><00:12:10.520><c> and</c><00:12:10.640><c> writers</c><00:12:11.080><c> suing</c><00:12:11.440><c> these</c>

00:12:11.630 --> 00:12:11.640 align:start position:0%
can. Artists and writers suing these
 

00:12:11.640 --> 00:12:13.430 align:start position:0%
can. Artists and writers suing these
companies<00:12:12.080><c> over</c><00:12:12.240><c> their</c><00:12:12.400><c> work</c><00:12:12.760><c> are,</c><00:12:13.120><c> in</c><00:12:13.240><c> her</c>

00:12:13.430 --> 00:12:13.440 align:start position:0%
companies over their work are, in her
 

00:12:13.440 --> 00:12:15.590 align:start position:0%
companies over their work are, in her
words,<00:12:14.120><c> trying</c><00:12:14.440><c> to</c><00:12:14.520><c> withhold</c><00:12:15.040><c> the</c><00:12:15.160><c> resource</c>

00:12:15.590 --> 00:12:15.600 align:start position:0%
words, trying to withhold the resource
 

00:12:15.600 --> 00:12:17.350 align:start position:0%
words, trying to withhold the resource
these<00:12:15.800><c> companies</c><00:12:16.200><c> need</c><00:12:16.400><c> most,</c><00:12:16.720><c> original</c>

00:12:17.350 --> 00:12:17.360 align:start position:0%
these companies need most, original
 

00:12:17.360 --> 00:12:19.870 align:start position:0%
these companies need most, original
human<00:12:17.720><c> data.</c><00:12:18.600><c> Get</c><00:12:18.720><c> loud</c><00:12:18.960><c> locally.</c>

00:12:19.870 --> 00:12:19.880 align:start position:0%
human data. Get loud locally.
 

00:12:19.880 --> 00:12:22.230 align:start position:0%
human data. Get loud locally.
&gt;&gt; And<00:12:20.080><c> that's</c><00:12:20.440><c> what</c><00:12:20.600><c> we've</c><00:12:20.800><c> seen</c><00:12:21.560><c> with</c><00:12:21.960><c> all</c><00:12:22.200><c> of</c>

00:12:22.230 --> 00:12:22.240 align:start position:0%
&gt;&gt; And that's what we've seen with all of
 

00:12:22.240 --> 00:12:23.630 align:start position:0%
&gt;&gt; And that's what we've seen with all of
these<00:12:22.440><c> people</c><00:12:22.760><c> that</c><00:12:22.880><c> are</c><00:12:22.960><c> now</c><00:12:23.200><c> literally</c>

00:12:23.630 --> 00:12:23.640 align:start position:0%
these people that are now literally
 

00:12:23.640 --> 00:12:24.830 align:start position:0%
these people that are now literally
protesting<00:12:24.080><c> the</c><00:12:24.160><c> streets</c><00:12:24.440><c> against</c><00:12:24.640><c> data</c>

00:12:24.830 --> 00:12:24.840 align:start position:0%
protesting the streets against data
 

00:12:24.840 --> 00:12:26.990 align:start position:0%
protesting the streets against data
centers<00:12:25.440><c> and</c><00:12:25.600><c> having</c><00:12:25.920><c> an</c><00:12:26.040><c> enormous</c><00:12:26.600><c> effect,</c>

00:12:26.990 --> 00:12:27.000 align:start position:0%
centers and having an enormous effect,
 

00:12:27.000 --> 00:12:29.270 align:start position:0%
centers and having an enormous effect,
by<00:12:27.120><c> the</c><00:12:27.240><c> way,</c><00:12:27.560><c> actually</c><00:12:28.040><c> stalling</c><00:12:29.040><c> data</c>

00:12:29.270 --> 00:12:29.280 align:start position:0%
by the way, actually stalling data
 

00:12:29.280 --> 00:12:31.430 align:start position:0%
by the way, actually stalling data
center<00:12:29.480><c> projects</c><00:12:30.200><c> and</c><00:12:30.320><c> also</c><00:12:30.960><c> completely</c>

00:12:31.430 --> 00:12:31.440 align:start position:0%
center projects and also completely
 

00:12:31.440 --> 00:12:33.030 align:start position:0%
center projects and also completely
banning<00:12:32.240><c> data</c><00:12:32.480><c> centers</c><00:12:32.760><c> from</c><00:12:32.880><c> being</c>

00:12:33.030 --> 00:12:33.040 align:start position:0%
banning data centers from being
 

00:12:33.040 --> 00:12:34.790 align:start position:0%
banning data centers from being
developed<00:12:33.400><c> in</c><00:12:33.520><c> their</c><00:12:33.640><c> localities.</c>

00:12:34.790 --> 00:12:34.800 align:start position:0%
developed in their localities.
 

00:12:34.800 --> 00:12:36.310 align:start position:0%
developed in their localities.
&gt;&gt; She<00:12:34.920><c> points</c><00:12:35.240><c> out</c><00:12:35.400><c> that</c><00:12:35.600><c> data</c><00:12:35.839><c> centers</c><00:12:36.240><c> are</c>

00:12:36.310 --> 00:12:36.320 align:start position:0%
&gt;&gt; She points out that data centers are
 

00:12:36.320 --> 00:12:38.630 align:start position:0%
&gt;&gt; She points out that data centers are
popping<00:12:36.760><c> up</c><00:12:36.880><c> in</c><00:12:37.080><c> ordinary</c><00:12:37.560><c> towns</c><00:12:38.000><c> right</c><00:12:38.200><c> now,</c>

00:12:38.630 --> 00:12:38.640 align:start position:0%
popping up in ordinary towns right now,
 

00:12:38.640 --> 00:12:40.829 align:start position:0%
popping up in ordinary towns right now,
and<00:12:38.760><c> in</c><00:12:38.880><c> several</c><00:12:39.280><c> places,</c><00:12:39.960><c> local</c><00:12:40.320><c> communities</c>

00:12:40.829 --> 00:12:40.839 align:start position:0%
and in several places, local communities
 

00:12:40.839 --> 00:12:42.750 align:start position:0%
and in several places, local communities
have<00:12:41.000><c> already</c><00:12:41.320><c> succeeded</c><00:12:41.839><c> in</c><00:12:41.960><c> stalling</c><00:12:42.560><c> or</c>

00:12:42.750 --> 00:12:42.760 align:start position:0%
have already succeeded in stalling or
 

00:12:42.760 --> 00:12:44.270 align:start position:0%
have already succeeded in stalling or
outright<00:12:43.200><c> banning</c><00:12:43.600><c> new</c><00:12:43.800><c> data</c><00:12:44.040><c> center</c>

00:12:44.270 --> 00:12:44.280 align:start position:0%
outright banning new data center
 

00:12:44.280 --> 00:12:46.670 align:start position:0%
outright banning new data center
projects<00:12:44.920><c> through</c><00:12:45.200><c> public</c><00:12:45.560><c> pressure.</c><00:12:46.480><c> Push</c>

00:12:46.670 --> 00:12:46.680 align:start position:0%
projects through public pressure. Push
 

00:12:46.680 --> 00:12:48.750 align:start position:0%
projects through public pressure. Push
at<00:12:46.760><c> your</c><00:12:46.880><c> school</c><00:12:47.360><c> or</c><00:12:47.480><c> workplace.</c>

00:12:48.750 --> 00:12:48.760 align:start position:0%
at your school or workplace.
 

00:12:48.760 --> 00:12:50.270 align:start position:0%
at your school or workplace.
If<00:12:48.839><c> your</c><00:12:49.000><c> company</c><00:12:49.440><c> or</c><00:12:49.560><c> university</c><00:12:50.160><c> is</c>

00:12:50.270 --> 00:12:50.280 align:start position:0%
If your company or university is
 

00:12:50.280 --> 00:12:52.910 align:start position:0%
If your company or university is
drafting<00:12:50.800><c> an</c><00:12:51.000><c> AI</c><00:12:51.320><c> adoption</c><00:12:51.839><c> policy,</c><00:12:52.680><c> that</c>

00:12:52.910 --> 00:12:52.920 align:start position:0%
drafting an AI adoption policy, that
 

00:12:52.920 --> 00:12:54.829 align:start position:0%
drafting an AI adoption policy, that
conversation<00:12:53.520><c> is</c><00:12:53.720><c> one</c><00:12:54.000><c> small</c><00:12:54.360><c> piece</c><00:12:54.680><c> of</c><00:12:54.760><c> a</c>

00:12:54.829 --> 00:12:54.839 align:start position:0%
conversation is one small piece of a
 

00:12:54.839 --> 00:12:56.710 align:start position:0%
conversation is one small piece of a
much<00:12:55.120><c> bigger</c><00:12:55.400><c> puzzle,</c><00:12:56.040><c> and</c><00:12:56.160><c> your</c><00:12:56.320><c> voice</c><00:12:56.640><c> in</c>

00:12:56.710 --> 00:12:56.720 align:start position:0%
much bigger puzzle, and your voice in
 

00:12:56.720 --> 00:12:59.070 align:start position:0%
much bigger puzzle, and your voice in
that<00:12:56.920><c> room</c><00:12:57.160><c> genuinely</c><00:12:57.720><c> counts.</c><00:12:58.640><c> Support</c>

00:12:59.070 --> 00:12:59.080 align:start position:0%
that room genuinely counts. Support
 

00:12:59.080 --> 00:13:00.510 align:start position:0%
that room genuinely counts. Support
alternatives.

00:13:00.510 --> 00:13:00.520 align:start position:0%
alternatives.
 

00:13:00.520 --> 00:13:02.390 align:start position:0%
alternatives.
Karen<00:13:00.839><c> says</c><00:13:01.440><c> there's</c><00:13:01.600><c> research</c><00:13:02.040><c> showing</c><00:13:02.280><c> the</c>

00:13:02.390 --> 00:13:02.400 align:start position:0%
Karen says there's research showing the
 

00:13:02.400 --> 00:13:05.190 align:start position:0%
Karen says there's research showing the
same<00:13:02.760><c> useful</c><00:13:03.240><c> AI</c><00:13:03.600><c> capabilities</c><00:13:04.640><c> can</c><00:13:04.760><c> be</c><00:13:04.920><c> built</c>

00:13:05.190 --> 00:13:05.200 align:start position:0%
same useful AI capabilities can be built
 

00:13:05.200 --> 00:13:07.350 align:start position:0%
same useful AI capabilities can be built
with<00:13:05.440><c> far</c><00:13:05.680><c> more</c><00:13:05.880><c> efficient</c><00:13:06.360><c> methods</c><00:13:06.920><c> and</c><00:13:07.120><c> far</c>

00:13:07.350 --> 00:13:07.360 align:start position:0%
with far more efficient methods and far
 

00:13:07.360 --> 00:13:09.390 align:start position:0%
with far more efficient methods and far
less<00:13:07.640><c> resource</c><00:13:08.080><c> consumption,</c><00:13:09.120><c> what</c><00:13:09.280><c> she</c>

00:13:09.390 --> 00:13:09.400 align:start position:0%
less resource consumption, what she
 

00:13:09.400 --> 00:13:11.790 align:start position:0%
less resource consumption, what she
calls<00:13:09.640><c> the</c><00:13:09.760><c> bicycles</c><00:13:10.400><c> of</c><00:13:10.520><c> AI</c><00:13:11.240><c> as</c><00:13:11.400><c> opposed</c><00:13:11.720><c> to</c>

00:13:11.790 --> 00:13:11.800 align:start position:0%
calls the bicycles of AI as opposed to
 

00:13:11.800 --> 00:13:14.310 align:start position:0%
calls the bicycles of AI as opposed to
the<00:13:11.920><c> resource-hungry</c><00:13:12.760><c> rockets.</c>

00:13:14.310 --> 00:13:14.320 align:start position:0%
the resource-hungry rockets.
 

00:13:14.320 --> 00:13:16.510 align:start position:0%
the resource-hungry rockets.
She<00:13:14.440><c> also</c><00:13:14.720><c> cites</c><00:13:15.000><c> a</c><00:13:15.080><c> striking</c><00:13:15.600><c> number.</c>

00:13:16.510 --> 00:13:16.520 align:start position:0%
She also cites a striking number.
 

00:13:16.520 --> 00:13:18.910 align:start position:0%
She also cites a striking number.
&gt;&gt; 80%<00:13:17.240><c> of</c><00:13:17.360><c> Americans</c><00:13:18.080><c> in</c><00:13:18.160><c> the</c><00:13:18.240><c> most</c><00:13:18.440><c> recent</c><00:13:18.720><c> poll</c>

00:13:18.910 --> 00:13:18.920 align:start position:0%
&gt;&gt; 80% of Americans in the most recent poll
 

00:13:18.920 --> 00:13:20.190 align:start position:0%
&gt;&gt; 80% of Americans in the most recent poll
think<00:13:19.120><c> that</c><00:13:19.240><c> the</c><00:13:19.360><c> AI</c><00:13:19.520><c> industry</c><00:13:19.839><c> need</c><00:13:20.000><c> to</c><00:13:20.080><c> be</c>

00:13:20.190 --> 00:13:20.200 align:start position:0%
think that the AI industry need to be
 

00:13:20.200 --> 00:13:21.590 align:start position:0%
think that the AI industry need to be
regulated.

00:13:21.590 --> 00:13:21.600 align:start position:0%
regulated.
 

00:13:21.600 --> 00:13:21.950 align:start position:0%
regulated.
&gt;&gt; Yeah,<00:13:21.839><c> I</c><00:13:21.880><c> see.</c>

00:13:21.950 --> 00:13:21.960 align:start position:0%
&gt;&gt; Yeah, I see.
 

00:13:21.960 --> 00:13:23.870 align:start position:0%
&gt;&gt; Yeah, I see.
&gt;&gt; When<00:13:22.120><c> was</c><00:13:22.320><c> the</c><00:13:22.440><c> last</c><00:13:22.839><c> time</c><00:13:23.160><c> that</c><00:13:23.320><c> 80%</c><00:13:23.800><c> of</c>

00:13:23.870 --> 00:13:23.880 align:start position:0%
&gt;&gt; When was the last time that 80% of
 

00:13:23.880 --> 00:13:25.230 align:start position:0%
&gt;&gt; When was the last time that 80% of
Americans<00:13:24.360><c> were</c><00:13:24.440><c> on</c><00:13:24.560><c> the</c><00:13:24.640><c> same</c><00:13:24.839><c> side</c><00:13:25.040><c> of</c><00:13:25.160><c> an</c>

00:13:25.230 --> 00:13:25.240 align:start position:0%
Americans were on the same side of an
 

00:13:25.240 --> 00:13:25.790 align:start position:0%
Americans were on the same side of an
issue?

00:13:25.790 --> 00:13:25.800 align:start position:0%
issue?
 

00:13:25.800 --> 00:13:27.430 align:start position:0%
issue?
&gt;&gt; A<00:13:25.880><c> shift</c><00:13:26.240><c> she</c><00:13:26.360><c> says</c><00:13:26.640><c> happened</c><00:13:27.000><c> largely</c>

00:13:27.430 --> 00:13:27.440 align:start position:0%
&gt;&gt; A shift she says happened largely
 

00:13:27.440 --> 00:13:28.990 align:start position:0%
&gt;&gt; A shift she says happened largely
because<00:13:27.760><c> of</c><00:13:27.840><c> stories</c><00:13:28.240><c> like</c><00:13:28.400><c> the</c><00:13:28.480><c> ones</c><00:13:28.760><c> in</c><00:13:28.840><c> her</c>

00:13:28.990 --> 00:13:29.000 align:start position:0%
because of stories like the ones in her
 

00:13:29.000 --> 00:13:31.510 align:start position:0%
because of stories like the ones in her
book<00:13:29.440><c> reaching</c><00:13:29.760><c> the</c><00:13:29.880><c> public.</c><00:13:30.920><c> I</c><00:13:31.000><c> want</c><00:13:31.240><c> to</c><00:13:31.360><c> end</c>

00:13:31.510 --> 00:13:31.520 align:start position:0%
book reaching the public. I want to end
 

00:13:31.520 --> 00:13:34.270 align:start position:0%
book reaching the public. I want to end
this<00:13:31.720><c> fairly</c><00:13:32.320><c> because</c><00:13:32.560><c> Karen</c><00:13:32.960><c> herself</c><00:13:33.520><c> does.</c>

00:13:34.270 --> 00:13:34.280 align:start position:0%
this fairly because Karen herself does.
 

00:13:34.280 --> 00:13:36.270 align:start position:0%
this fairly because Karen herself does.
Late<00:13:34.560><c> in</c><00:13:34.640><c> the</c><00:13:34.720><c> conversation,</c><00:13:35.600><c> Steven</c><00:13:36.000><c> pushes</c>

00:13:36.270 --> 00:13:36.280 align:start position:0%
Late in the conversation, Steven pushes
 

00:13:36.280 --> 00:13:38.190 align:start position:0%
Late in the conversation, Steven pushes
back<00:13:36.560><c> a</c><00:13:36.600><c> little</c><00:13:37.040><c> saying</c><00:13:37.360><c> that</c><00:13:37.800><c> as</c><00:13:38.000><c> an</c>

00:13:38.190 --> 00:13:38.200 align:start position:0%
back a little saying that as an
 

00:13:38.200 --> 00:13:40.590 align:start position:0%
back a little saying that as an
entrepreneur<00:13:38.960><c> and</c><00:13:39.200><c> investor,</c><00:13:39.880><c> he</c><00:13:40.040><c> genuinely</c>

00:13:40.590 --> 00:13:40.600 align:start position:0%
entrepreneur and investor, he genuinely
 

00:13:40.600 --> 00:13:43.590 align:start position:0%
entrepreneur and investor, he genuinely
finds<00:13:41.040><c> AI</c><00:13:41.560><c> incredible.</c><00:13:42.520><c> Safer</c><00:13:42.920><c> driving,</c>

00:13:43.590 --> 00:13:43.600 align:start position:0%
finds AI incredible. Safer driving,
 

00:13:43.600 --> 00:13:45.270 align:start position:0%
finds AI incredible. Safer driving,
tools<00:13:43.920><c> that</c><00:13:44.040><c> help</c><00:13:44.240><c> him</c><00:13:44.400><c> create</c><00:13:44.800><c> things</c><00:13:45.120><c> he</c>

00:13:45.270 --> 00:13:45.280 align:start position:0%
tools that help him create things he
 

00:13:45.280 --> 00:13:47.590 align:start position:0%
tools that help him create things he
couldn't<00:13:45.560><c> before.</c><00:13:46.320><c> Real</c><00:13:46.640><c> value</c><00:13:47.080><c> for</c><00:13:47.280><c> over</c><00:13:47.560><c> a</c>

00:13:47.590 --> 00:13:47.600 align:start position:0%
couldn't before. Real value for over a
 

00:13:47.600 --> 00:13:49.990 align:start position:0%
couldn't before. Real value for over a
billion<00:13:48.040><c> users.</c><00:13:48.960><c> And</c><00:13:49.080><c> Karen</c><00:13:49.360><c> doesn't</c><00:13:49.640><c> dismiss</c>

00:13:49.990 --> 00:13:50.000 align:start position:0%
billion users. And Karen doesn't dismiss
 

00:13:50.000 --> 00:13:52.390 align:start position:0%
billion users. And Karen doesn't dismiss
that.<00:13:50.640><c> Her</c><00:13:50.840><c> answer</c><00:13:51.120><c> is</c><00:13:51.240><c> basically,</c><00:13:52.160><c> you</c><00:13:52.280><c> can</c>

00:13:52.390 --> 00:13:52.400 align:start position:0%
that. Her answer is basically, you can
 

00:13:52.400 --> 00:13:54.710 align:start position:0%
that. Her answer is basically, you can
hold<00:13:52.680><c> both</c><00:13:53.000><c> truths</c><00:13:53.360><c> at</c><00:13:53.520><c> once.</c>

00:13:54.710 --> 00:13:54.720 align:start position:0%
hold both truths at once.
 

00:13:54.720 --> 00:13:56.830 align:start position:0%
hold both truths at once.
AI<00:13:55.000><c> can</c><00:13:55.160><c> be</c><00:13:55.280><c> genuinely</c><00:13:55.920><c> useful</c><00:13:56.640><c> and</c><00:13:56.760><c> the</c>

00:13:56.830 --> 00:13:56.840 align:start position:0%
AI can be genuinely useful and the
 

00:13:56.840 --> 00:13:58.590 align:start position:0%
AI can be genuinely useful and the
current<00:13:57.160><c> way</c><00:13:57.280><c> it's</c><00:13:57.440><c> being</c><00:13:57.640><c> built</c><00:13:58.080><c> and</c><00:13:58.280><c> rushed</c>

00:13:58.590 --> 00:13:58.600 align:start position:0%
current way it's being built and rushed
 

00:13:58.600 --> 00:14:01.310 align:start position:0%
current way it's being built and rushed
to<00:13:58.680><c> market</c><00:13:59.400><c> can</c><00:13:59.560><c> be</c><00:13:59.720><c> genuinely</c><00:14:00.240><c> harmful.</c>

00:14:01.310 --> 00:14:01.320 align:start position:0%
to market can be genuinely harmful.
 

00:14:01.320 --> 00:14:03.310 align:start position:0%
to market can be genuinely harmful.
Those<00:14:01.480><c> aren't</c><00:14:01.680><c> contradictions.</c><00:14:02.680><c> They're</c><00:14:02.880><c> two</c>

00:14:03.310 --> 00:14:03.320 align:start position:0%
Those aren't contradictions. They're two
 

00:14:03.320 --> 00:14:05.150 align:start position:0%
Those aren't contradictions. They're two
honest<00:14:03.760><c> observations</c><00:14:04.520><c> about</c><00:14:04.720><c> the</c><00:14:04.840><c> same</c>

00:14:05.150 --> 00:14:05.160 align:start position:0%
honest observations about the same
 

00:14:05.160 --> 00:14:07.110 align:start position:0%
honest observations about the same
industry.<00:14:06.080><c> That's</c><00:14:06.360><c> actually</c><00:14:06.720><c> the</c><00:14:06.800><c> most</c>

00:14:07.110 --> 00:14:07.120 align:start position:0%
industry. That's actually the most
 

00:14:07.120 --> 00:14:10.310 align:start position:0%
industry. That's actually the most
balanced<00:14:08.040><c> moment</c><00:14:08.440><c> in</c><00:14:08.520><c> the</c><00:14:08.640><c> whole</c><00:14:09.080><c> two</c><00:14:09.360><c> hours.</c>

00:14:10.310 --> 00:14:10.320 align:start position:0%
balanced moment in the whole two hours.
 

00:14:10.320 --> 00:14:12.550 align:start position:0%
balanced moment in the whole two hours.
And<00:14:10.480><c> it's</c><00:14:10.640><c> the</c><00:14:10.760><c> note</c><00:14:11.000><c> I</c><00:14:11.040><c> want</c><00:14:11.200><c> to</c><00:14:11.280><c> end</c><00:14:11.440><c> on,</c><00:14:11.600><c> too.</c>

00:14:12.550 --> 00:14:12.560 align:start position:0%
And it's the note I want to end on, too.
 

00:14:12.560 --> 00:14:14.630 align:start position:0%
And it's the note I want to end on, too.
So,<00:14:12.680><c> here's</c><00:14:12.960><c> my</c><00:14:13.120><c> question</c><00:14:13.560><c> for</c><00:14:13.760><c> you</c><00:14:14.440><c> and</c><00:14:14.560><c> I</c>

00:14:14.630 --> 00:14:14.640 align:start position:0%
So, here's my question for you and I
 

00:14:14.640 --> 00:14:16.830 align:start position:0%
So, here's my question for you and I
genuinely<00:14:15.160><c> want</c><00:14:15.320><c> to</c><00:14:15.400><c> know.</c><00:14:16.120><c> Knowing</c><00:14:16.560><c> what</c><00:14:16.720><c> you</c>

00:14:16.830 --> 00:14:16.840 align:start position:0%
genuinely want to know. Knowing what you
 

00:14:16.840 --> 00:14:18.470 align:start position:0%
genuinely want to know. Knowing what you
now<00:14:17.080><c> know</c><00:14:17.280><c> about</c><00:14:17.560><c> where</c><00:14:17.760><c> the</c><00:14:17.880><c> water</c><00:14:18.240><c> for</c><00:14:18.360><c> your</c>

00:14:18.470 --> 00:14:18.480 align:start position:0%
now know about where the water for your
 

00:14:18.480 --> 00:14:20.270 align:start position:0%
now know about where the water for your
chatbot's<00:14:19.080><c> training</c><00:14:19.480><c> run</c><00:14:19.640><c> might</c><00:14:19.800><c> be</c><00:14:19.920><c> coming</c>

00:14:20.270 --> 00:14:20.280 align:start position:0%
chatbot's training run might be coming
 

00:14:20.280 --> 00:14:22.310 align:start position:0%
chatbot's training run might be coming
from,<00:14:21.080><c> does</c><00:14:21.280><c> it</c><00:14:21.400><c> change</c><00:14:21.760><c> how</c><00:14:21.920><c> you'll</c><00:14:22.080><c> use</c>

00:14:22.310 --> 00:14:22.320 align:start position:0%
from, does it change how you'll use
 

00:14:22.320 --> 00:14:24.750 align:start position:0%
from, does it change how you'll use
these<00:14:22.560><c> tools</c><00:14:22.800><c> tomorrow?</c><00:14:23.840><c> Not</c><00:14:24.000><c> never.</c><00:14:24.440><c> I'm</c><00:14:24.560><c> not</c>

00:14:24.750 --> 00:14:24.760 align:start position:0%
these tools tomorrow? Not never. I'm not
 

00:14:24.760 --> 00:14:27.670 align:start position:0%
these tools tomorrow? Not never. I'm not
asking<00:14:25.040><c> you</c><00:14:25.120><c> to</c><00:14:25.240><c> swear</c><00:14:25.520><c> off</c><00:14:25.640><c> technology.</c><00:14:26.760><c> Just</c>

00:14:27.670 --> 00:14:27.680 align:start position:0%
asking you to swear off technology. Just
 

00:14:27.680 --> 00:14:30.310 align:start position:0%
asking you to swear off technology. Just
pay<00:14:27.800><c> attention.</c><00:14:28.760><c> Ask</c><00:14:29.040><c> questions.</c>

00:14:30.310 --> 00:14:30.320 align:start position:0%
pay attention. Ask questions.
 

00:14:30.320 --> 00:14:32.510 align:start position:0%
pay attention. Ask questions.
Support<00:14:30.720><c> the</c><00:14:30.800><c> version</c><00:14:31.160><c> of</c><00:14:31.320><c> AI</c><00:14:32.080><c> that</c><00:14:32.200><c> doesn't</c>

00:14:32.510 --> 00:14:32.520 align:start position:0%
Support the version of AI that doesn't
 

00:14:32.520 --> 00:14:34.350 align:start position:0%
Support the version of AI that doesn't
need<00:14:32.720><c> a</c><00:14:32.760><c> small</c><00:14:33.080><c> town</c><00:14:33.360><c> in</c><00:14:33.440><c> Tennessee</c><00:14:33.920><c> to</c><00:14:34.040><c> smell</c>

00:14:34.350 --> 00:14:34.360 align:start position:0%
need a small town in Tennessee to smell
 

00:14:34.360 --> 00:14:35.950 align:start position:0%
need a small town in Tennessee to smell
gas<00:14:34.880><c> in</c><00:14:35.000><c> their</c><00:14:35.120><c> living</c><00:14:35.400><c> room</c><00:14:35.560><c> to</c><00:14:35.680><c> make</c><00:14:35.839><c> it</c>

00:14:35.950 --> 00:14:35.960 align:start position:0%
gas in their living room to make it
 

00:14:35.960 --> 00:14:37.590 align:start position:0%
gas in their living room to make it
work.<00:14:36.680><c> If</c><00:14:36.839><c> this</c><00:14:37.040><c> made</c><00:14:37.240><c> you</c><00:14:37.360><c> think</c>

00:14:37.590 --> 00:14:37.600 align:start position:0%
work. If this made you think
 

00:14:37.600 --> 00:14:39.990 align:start position:0%
work. If this made you think
differently,<00:14:38.440><c> even</c><00:14:38.680><c> a</c><00:14:38.720><c> little,</c><00:14:39.520><c> share</c><00:14:39.880><c> it</c>

00:14:39.990 --> 00:14:40.000 align:start position:0%
differently, even a little, share it
 

00:14:40.000 --> 00:14:42.270 align:start position:0%
differently, even a little, share it
with<00:14:40.160><c> one</c><00:14:40.320><c> person</c><00:14:40.800><c> who</c><00:14:40.960><c> uses</c><00:14:41.360><c> AI</c><00:14:41.760><c> every</c><00:14:42.080><c> day</c>

00:14:42.270 --> 00:14:42.280 align:start position:0%
with one person who uses AI every day
 

00:14:42.280 --> 00:14:44.430 align:start position:0%
with one person who uses AI every day
without<00:14:42.640><c> ever</c><00:14:42.880><c> asking</c><00:14:43.200><c> these</c><00:14:43.360><c> questions.</c>

00:14:44.430 --> 00:14:44.440 align:start position:0%
without ever asking these questions.
 

00:14:44.440 --> 00:14:46.190 align:start position:0%
without ever asking these questions.
That's<00:14:44.760><c> genuinely</c><00:14:45.320><c> how</c><00:14:45.480><c> change</c><00:14:45.839><c> like</c><00:14:46.000><c> this</c>

00:14:46.190 --> 00:14:46.200 align:start position:0%
That's genuinely how change like this
 

00:14:46.200 --> 00:14:48.800 align:start position:0%
That's genuinely how change like this
starts.