Real Coffee with Scott Adams - Episode 2084 Scott Adams： Starship Launch, Hunter's Whistleblower, AI Bias, Can AI Be Hypnotized？

00:00:00.000 Doo Doo Doo Doo Doo Doo Doo Doo Doo Doo Doo Doo 0.95

00:00:03.340 Doo Doo Doo Doo Doo Doo 0.91

00:00:05.240 Good morning everybody and welcome to the Highlight of civilization.

00:00:11.040 And it really is.

00:00:12.380 Today does feel like the highlight of civilization.

00:00:15.220 It's a great day.

00:00:16.820 I'm so happy.

00:00:18.180 And I think you will be too.

00:00:19.900 Because by the time you get done with this live stream,

00:00:23.220 oh your dopamine is going to be flowing.

00:00:26.060 Your oxytocin will be off the charts.

00:00:28.240 If you need a little testosterone, I got it.

00:00:32.260 I got it here for you.

00:00:33.880 And all you need to get all of those benefits

00:00:36.060 is a cup or a mug or a glass,

00:00:38.320 a tanker, a tanker, a chalice, a stein,

00:00:41.220 a canteen jug or flask, a vessel of any kind.

00:00:44.580 Fill it with your favorite liquid.

00:00:47.000 I like coffee.

00:00:48.520 And join me now for some un...

00:00:51.240 Join me now for the simultaneous sip

00:00:55.440 or something like that.

00:00:56.820 Go.

00:00:58.240 Pause. Pause.

00:01:04.820 Don't put down your cup.

00:01:07.120 Do not put down your cup.

00:01:09.400 Today is a special day.

00:01:11.400 How many of you got to watch the launch of the Starship?

00:01:16.420 It just happened.

00:01:17.860 Now, if you didn't see it,

00:01:19.940 the first stage of the test was successful.

00:01:22.880 It launched.

00:01:24.480 They did hope that they would get a separation

00:01:27.060 and then the second part of it would do a little bit of an orbit.

00:01:31.020 But that part did not work.

00:01:34.820 However,

00:01:35.940 however,

00:01:38.340 I'm pretty sure that Elon Musk is a little bit disappointed

00:01:44.680 because it didn't do both parts of the test correctly.

00:01:50.140 However,

00:01:51.140 the first part of the test was very important

00:01:53.800 because it got them data that will help them with the second part of the test,

00:01:56.940 which they will do.

00:01:57.960 And here's what I would like to add to this moment.

00:02:02.940 I can't remember ever being more excited than I was when I was watching the takeoff.

00:02:11.260 Did anybody have that feeling?

00:02:12.400 I don't think I've ever been more excited watching something on television.

00:02:18.320 I've watched the moonwalk live.

00:02:23.780 I'm the age where I got to watch the moonwalk.

00:02:27.500 But I've never been more excited than that launch.

00:02:32.980 And unlike,

00:02:35.540 you know,

00:02:36.220 maybe Elon Musk might be a little disappointed

00:02:39.400 he didn't do the second part of the test

00:02:41.440 or didn't do the second part successfully.

00:02:43.440 But I would like to say that

00:02:46.180 what we love about Elon Musk

00:02:48.860 and maybe about America too

00:02:51.160 is

00:02:52.960 not that it

00:02:54.780 the big story

00:02:56.940 let me just make sure I can frame the story right.

00:03:01.680 The news will be talking about the launch all day long.

00:03:04.540 Let me frame it for you.

00:03:06.120 This is the way you should see it.

00:03:08.240 The amazing thing

00:03:09.740 is not the launch

00:03:11.100 and it's not the fact that it didn't completely succeed.

00:03:15.480 You know what the amazing thing was?

00:03:18.180 That Elon Musk knew

00:03:19.540 there was a good, good chance

00:03:21.940 it was going to blow up

00:03:23.020 and he did it anyway.

00:03:26.320 He did it anyway.

00:03:29.020 That's everything.

00:03:30.840 That's everything.

00:03:32.460 The story is not about the technology

00:03:34.300 which was amazing.

00:03:36.100 The story is not that

00:03:37.120 half of it worked and half of it didn't.

00:03:39.500 That's just the detail.

00:03:41.100 The amazing part of the story

00:03:43.080 is that he knew

00:03:45.000 it would probably blow up 0.99

00:03:46.320 in front of the whole damn world 0.99

00:03:48.180 and he did it anyway. 0.99

00:03:50.820 Wow.

00:03:52.240 Here's him.

00:03:53.020 That was inspiring.

00:04:06.800 Are you ready for the best 420 story

00:04:09.320 of all time?

00:04:10.920 Now I don't know if this is true

00:04:12.660 and it doesn't matter.

00:04:15.980 Does it?

00:04:16.740 It probably won't matter at all

00:04:18.220 if it's true.

00:04:18.740 But you know actor

00:04:21.020 Woody Harrelson

00:04:25.240 and you know

00:04:27.720 McConaughey, right?

00:04:29.640 What's his name?

00:04:30.840 Matthew McConaughey.

00:04:32.840 So just think about

00:04:34.080 both of those actors

00:04:35.100 for a moment.

00:04:36.720 Hold them in your mind.

00:04:37.960 Woody Harrelson

00:04:38.860 Woody Harrelson

00:04:40.780 and Matthew McConaughey.

00:04:42.980 Well it turns out

00:04:44.840 that because they're both

00:04:46.120 famous actors

00:04:46.940 and they're also

00:04:47.820 going to be in some kind of

00:04:49.460 project upcoming.

00:04:52.360 True detective

00:04:53.060 is that what it is?

00:04:54.240 So they know each other

00:04:55.480 and they're

00:04:56.040 good friends.

00:04:58.560 You know

00:04:58.820 off screen

00:04:59.380 they're friends.

00:05:00.360 And one day

00:05:00.820 they were hanging out

00:05:01.580 at some sporting event

00:05:03.720 and Matthew McConaughey's

00:05:05.880 mother was there.

00:05:06.560 And

00:05:08.740 apparently

00:05:10.380 she let slip

00:05:11.500 a little tidbit

00:05:13.440 that makes it possible

00:05:15.540 that she is

00:05:17.540 the mother

00:05:17.900 of both of them.

00:05:22.780 So

00:05:23.380 apparently

00:05:24.420 they might be brothers.

00:05:26.840 Now it's

00:05:27.600 unconfirmed

00:05:28.480 but they both

00:05:29.740 see it

00:05:30.200 like once

00:05:31.380 it became

00:05:32.940 a possibility.

00:05:34.740 I think

00:05:34.920 and you see

00:05:35.920 the picture

00:05:36.540 of them

00:05:36.820 standing

00:05:37.140 next to each other

00:05:38.060 and as soon

00:05:39.400 as you see

00:05:39.900 them standing

00:05:40.400 next to each other

00:05:41.300 and then you think

00:05:42.480 about both

00:05:43.020 of their personalities

00:05:43.960 and you know

00:05:45.560 that apparently

00:05:46.340 their mother

00:05:47.080 knew both

00:05:49.220 of the fathers

00:05:49.880 and the timing

00:05:51.420 lined up

00:05:52.160 when one of them

00:05:53.400 was not with

00:05:54.300 the other father

00:05:55.560 and one was

00:05:56.860 in town

00:05:57.360 and they didn't

00:05:59.060 apparently

00:05:59.600 the mother

00:06:00.060 knew both

00:06:00.520 of the fathers.

00:06:02.040 Let's put it

00:06:02.480 this way.

00:06:03.180 The thing

00:06:03.580 that's confirmed

00:06:04.220 is that

00:06:05.220 the mom

00:06:05.940 knew

00:06:06.760 both

00:06:07.200 of their

00:06:07.480 fathers.

00:06:11.280 Anyway

00:06:11.920 that's just

00:06:13.780 your perfect

00:06:14.300 420 story

00:06:15.320 but it gets

00:06:16.260 better.

00:06:18.340 So

00:06:18.940 last night

00:06:19.740 I was

00:06:21.060 listening in

00:06:22.840 on one

00:06:23.560 of the great

00:06:24.100 spaces events

00:06:27.280 on Twitter

00:06:28.220 and wow

00:06:29.740 was it good.

00:06:30.740 It was

00:06:30.940 some of the

00:06:31.540 smartest people

00:06:32.200 I've ever heard

00:06:32.780 in my life

00:06:33.440 talking about

00:06:34.600 AI.

00:06:35.940 Now

00:06:36.260 when I say

00:06:37.540 smartest people

00:06:38.560 it's just

00:06:39.640 some of the

00:06:40.080 smartest people

00:06:40.660 I've ever

00:06:41.380 ever experienced

00:06:42.280 in the same

00:06:42.980 place

00:06:43.280 at the same

00:06:43.700 time.

00:06:44.600 It was

00:06:45.000 unbelievable

00:06:46.280 just having it

00:06:47.440 in my

00:06:47.880 earbuds

00:06:48.780 and

00:06:49.280 listening to

00:06:50.700 one person

00:06:51.360 after another

00:06:51.940 was just

00:06:52.420 freaking

00:06:53.020 brilliant.

00:06:54.580 And Brian

00:06:55.100 Romelli

00:06:55.560 I always

00:06:55.940 recommend him

00:06:56.700 he's the guy

00:06:58.080 you've got to

00:06:58.580 follow.

00:06:59.560 If you don't

00:07:00.280 want AI

00:07:00.800 to sneak up 0.89

00:07:01.720 and kick you 1.00

00:07:02.180 in the ass 1.00

00:07:02.680 you've got to 1.00

00:07:03.820 follow him

00:07:04.320 he's way ahead

00:07:05.420 of everything.

00:07:06.540 So

00:07:06.740 I'll tell you

00:07:08.640 some of the

00:07:08.940 things that

00:07:09.360 came out of

00:07:09.840 that.

00:07:11.120 Number one

00:07:11.880 here's my

00:07:13.360 conclusion

00:07:13.900 sort of

00:07:14.760 summarizing

00:07:15.500 or my

00:07:16.340 opinion based

00:07:16.880 on what I

00:07:17.240 heard.

00:07:17.420 number one

00:07:20.200 is AI

00:07:21.420 is never

00:07:21.920 going to be

00:07:22.340 unbiased.

00:07:23.840 Do you

00:07:24.160 know that?

00:07:25.740 It's not

00:07:26.460 even a

00:07:26.780 possibility

00:07:27.320 because it's

00:07:28.940 trained on

00:07:29.880 human patterns

00:07:31.380 of language.

00:07:32.440 They're called

00:07:33.260 LLMs or

00:07:34.660 large language

00:07:35.680 models.

00:07:36.680 It doesn't

00:07:37.460 think it

00:07:38.760 just looks for

00:07:39.380 the same

00:07:39.820 patterns that

00:07:41.360 would be

00:07:41.680 common to

00:07:42.180 human usage.

00:07:44.020 So I've

00:07:44.880 developed a

00:07:45.980 hypothesis

00:07:47.680 and I'm

00:07:49.220 going to

00:07:49.400 call it

00:07:49.720 the

00:07:49.940 Adam's

00:07:50.500 law of

00:07:50.960 AI.

00:07:52.400 You ready

00:07:53.080 for this?

00:07:53.940 So here's

00:07:54.340 the Adam's

00:07:54.960 law of

00:07:55.340 AI.

00:07:56.800 There's a

00:07:57.620 natural

00:07:58.040 limit to

00:08:00.880 how smart

00:08:02.260 AI can get

00:08:03.260 and that

00:08:04.200 natural limit

00:08:05.080 is the

00:08:06.080 smartest

00:08:06.400 human.

00:08:08.140 So AI

00:08:08.720 will be

00:08:09.320 way smarter

00:08:10.120 than almost

00:08:11.520 all humans

00:08:12.280 at almost

00:08:12.820 all things

00:08:13.520 but then it

00:08:14.820 can't get

00:08:15.360 smarter than

00:08:15.920 the smartest

00:08:16.380 human.

00:08:17.840 It can

00:08:18.540 give faster, 0.98

00:08:20.120 right?

00:08:20.780 Let's be

00:08:21.180 careful.

00:08:22.220 It can

00:08:22.580 give faster

00:08:23.420 which turns

00:08:24.860 out to be a

00:08:25.360 huge advantage

00:08:25.980 so it would

00:08:27.080 look like

00:08:27.520 intelligence

00:08:28.020 in its own

00:08:29.000 way.

00:08:29.700 But I don't

00:08:30.480 think it

00:08:30.920 can get

00:08:31.460 smarter.

00:08:33.160 And when I

00:08:33.600 say smarter

00:08:34.280 I don't mean

00:08:35.600 doing math

00:08:36.640 or maybe

00:08:37.420 searching

00:08:37.900 through things

00:08:38.520 quicker

00:08:38.920 which it

00:08:40.340 will do

00:08:40.580 very well.

00:08:41.760 I mean

00:08:42.340 it won't

00:08:42.840 be able to

00:08:43.280 understand

00:08:43.780 its environment

00:08:44.620 and interpret

00:08:45.960 it any

00:08:47.140 better than

00:08:47.660 the best

00:08:48.060 person can

00:08:48.700 do it

00:08:49.140 because it

00:08:50.140 won't have

00:08:50.480 any way

00:08:50.960 to do

00:08:51.240 that.

00:08:51.940 I don't

00:08:52.140 think you

00:08:52.520 can look

00:08:53.800 at language

00:08:54.360 of humans

00:08:55.000 and then

00:08:55.700 with that

00:08:56.740 alone

00:08:57.220 go to a

00:08:58.920 higher level

00:08:59.420 of intelligence.

00:09:00.540 It's just

00:09:01.160 going to be

00:09:01.460 the average

00:09:01.940 or the

00:09:02.660 preponderance

00:09:03.880 or the

00:09:04.220 consensus

00:09:04.740 or some

00:09:06.220 clever

00:09:06.880 algorithm

00:09:07.520 of just

00:09:08.140 people talk.

00:09:09.500 that's all

00:09:10.520 it can

00:09:10.740 do.

00:09:12.400 And so

00:09:12.900 one of the

00:09:14.320 questions you

00:09:14.840 might ask

00:09:15.200 yourself is

00:09:15.780 what subset

00:09:17.440 of human

00:09:18.360 activities or

00:09:19.280 conversations

00:09:20.000 is it using

00:09:21.200 to train

00:09:21.700 itself?

00:09:22.600 Well let me

00:09:23.020 tell you.

00:09:24.440 Wouldn't you

00:09:24.840 like to know?

00:09:26.040 It turns out

00:09:26.560 that the

00:09:26.980 Washington

00:09:28.140 Post

00:09:29.120 looked into

00:09:30.660 it and

00:09:31.180 tried to

00:09:31.760 figure out

00:09:32.300 where it

00:09:32.980 was getting

00:09:33.320 most of

00:09:33.760 its data.

00:09:35.100 and I

00:09:37.500 of course

00:09:37.920 wrote that

00:09:38.380 down and

00:09:38.820 can't find

00:09:39.340 it in my

00:09:39.700 notes but

00:09:40.120 the bottom

00:09:40.880 line is that

00:09:41.640 it's the

00:09:42.300 New York

00:09:43.620 Times,

00:09:44.740 Washington

00:09:45.080 Post,

00:09:46.060 here it

00:09:46.500 is.

00:09:47.680 So by

00:09:48.100 half of

00:09:48.500 the top

00:09:48.840 ten sites

00:09:49.620 this is

00:09:50.120 from

00:09:50.500 Washington

00:09:50.880 Post

00:09:51.140 reporting,

00:09:52.180 half of

00:09:52.680 the top

00:09:53.000 ten sites

00:09:53.660 overall

00:09:54.100 were news

00:09:54.660 outlets.

00:09:56.140 So the

00:09:57.680 biggest AI

00:09:59.000 that everybody

00:10:00.120 talks about

00:10:00.720 is training

00:10:01.680 mostly on

00:10:02.820 the big

00:10:03.660 news sites.

00:10:05.100 you know

00:10:05.880 that's a

00:10:06.340 not over

00:10:07.880 half but

00:10:08.400 it's just

00:10:08.700 the biggest

00:10:09.140 chunk.

00:10:10.700 So that

00:10:10.860 would include

00:10:11.300 New York

00:10:11.800 Times,

00:10:12.260 Washington

00:10:12.480 Post,

00:10:12.860 LA Times,

00:10:13.880 The Guardian,

00:10:15.260 Forbes,

00:10:16.220 and Huffington

00:10:16.780 Post.

00:10:18.560 Washington

00:10:19.080 Post was

00:10:19.540 number 11.

00:10:22.280 So the

00:10:22.880 Huffington

00:10:23.460 Post is

00:10:25.460 in the

00:10:26.120 top ten

00:10:27.060 news sites

00:10:28.800 that AI

00:10:29.780 is training

00:10:30.280 on.

00:10:32.120 Let me

00:10:32.880 say that

00:10:33.580 again.

00:10:35.100 we're

00:10:35.600 trying to

00:10:36.100 build

00:10:36.500 an

00:10:37.400 advanced

00:10:37.980 intelligence

00:10:38.820 with input

00:10:41.920 from the

00:10:42.300 Huffington

00:10:42.660 Post.

00:10:47.880 Okay.

00:10:51.600 Now,

00:10:52.700 they also

00:10:53.840 have,

00:10:54.440 they also

00:10:54.900 are pulling

00:10:55.360 material from

00:10:56.520 Russia today,

00:10:58.380 but not much.

00:10:59.740 That's like

00:11:00.140 65th on the

00:11:01.260 list.

00:11:01.520 that's a

00:11:03.000 Russian state

00:11:03.740 propaganda

00:11:04.760 site.

00:11:05.820 And then

00:11:06.040 because it's

00:11:06.560 the Washington

00:11:07.000 Post,

00:11:07.780 they try to

00:11:08.540 lump

00:11:08.760 Breitbart

00:11:09.240 into the

00:11:09.900 same paragraph

00:11:10.500 as Russia

00:11:11.480 today.

00:11:13.520 They're just 1.00

00:11:14.360 bastards. 1.00

00:11:15.340 They're just 1.00

00:11:15.700 freaking 1.00

00:11:15.980 bastards. 1.00

00:11:18.020 So the 1.00

00:11:18.840 Washington

00:11:19.720 Post,

00:11:20.400 this is

00:11:20.820 just a

00:11:21.620 perfect

00:11:21.840 example.

00:11:23.400 Joel,

00:11:23.720 if you're

00:11:23.960 watching this,

00:11:25.140 you'll enjoy

00:11:26.360 this.

00:11:26.640 So the

00:11:27.360 Washington

00:11:27.740 Post is

00:11:28.400 reporting how

00:11:29.260 AI is

00:11:29.900 gaining its

00:11:30.460 intelligence and

00:11:31.840 points out that

00:11:32.560 the Washington

00:11:33.040 Post itself is

00:11:34.980 the 11th biggest

00:11:36.300 source among the

00:11:37.540 news sources

00:11:38.560 that it's looking

00:11:39.500 at.

00:11:39.940 So that's

00:11:40.300 pretty good for

00:11:40.800 the Washington

00:11:41.200 Post,

00:11:41.660 right?

00:11:42.780 They're right up

00:11:43.720 there in

00:11:44.460 trading AI.

00:11:45.640 That's good.

00:11:46.560 Except in the

00:11:47.540 very same

00:11:48.360 Washington Post

00:11:49.440 article in which

00:11:50.160 they brag about

00:11:50.980 being one of the

00:11:51.760 sources,

00:11:52.540 they've got

00:11:53.280 blatant propaganda

00:11:54.660 by putting

00:11:55.940 Russia Today,

00:11:57.200 which is

00:11:57.720 literally owned

00:11:58.660 by Putin,

00:11:59.680 in the same

00:12:02.620 block like

00:12:03.760 they're lumping

00:12:04.940 them in the

00:12:05.360 same category,

00:12:06.200 Breitbart.

00:12:07.420 Okay.

00:12:10.080 Now,

00:12:11.080 do I have to

00:12:11.940 make my case

00:12:12.740 that AI will

00:12:13.880 be biased?

00:12:15.760 It's Fed

00:12:16.260 biased.

00:12:18.140 How could

00:12:18.720 it not be?

00:12:19.820 There's no

00:12:20.680 possibility it

00:12:21.440 could be

00:12:21.600 anything but

00:12:22.120 biased.

00:12:22.520 AI will

00:12:25.840 only be as

00:12:26.560 smart as

00:12:27.420 the people

00:12:27.960 who are

00:12:28.640 feeding it

00:12:29.100 information

00:12:29.720 decide what

00:12:31.240 database it

00:12:32.100 looks at.

00:12:33.220 And then it

00:12:33.600 can't get

00:12:34.020 smarter than

00:12:34.620 the smartest

00:12:35.120 person in

00:12:35.780 those databases.

00:12:37.420 How would

00:12:38.100 it do that?

00:12:39.260 It would

00:12:39.580 know more,

00:12:40.940 but a human

00:12:41.920 could know

00:12:42.380 more if it

00:12:42.980 took more

00:12:43.400 time.

00:12:43.820 So it's

00:12:44.040 really a

00:12:44.380 time thing

00:12:44.960 more than

00:12:45.840 a thinking

00:12:46.400 thing.

00:12:47.940 All right.

00:12:48.800 So just to

00:12:49.560 be clear,

00:12:50.100 of course AI

00:12:50.660 will be far

00:12:51.520 smarter than

00:12:51.980 humans,

00:12:52.520 in a whole

00:12:52.880 bunch of

00:12:53.200 ways,

00:12:53.660 and far

00:12:54.060 faster.

00:12:55.180 But in

00:12:55.700 the basic,

00:12:56.640 most important

00:12:57.920 part of

00:12:59.600 being a

00:13:00.060 human,

00:13:01.020 which is

00:13:01.620 what is

00:13:02.120 real in

00:13:02.620 my immediate

00:13:03.240 environment?

00:13:04.620 What are

00:13:05.200 these people

00:13:05.740 thinking?

00:13:07.460 How do I

00:13:08.280 act to get

00:13:09.160 a good result

00:13:09.860 in my

00:13:10.360 situation?

00:13:11.840 None of

00:13:12.340 that stuff

00:13:12.840 the AI is

00:13:13.460 going to

00:13:13.620 help you

00:13:13.880 with,

00:13:14.480 because humans

00:13:15.420 created it.

00:13:18.400 All right.

00:13:18.820 there was a

00:13:21.700 big discussion

00:13:22.300 in the

00:13:22.780 same

00:13:23.120 spaces

00:13:23.640 group

00:13:24.000 that I

00:13:24.340 attended

00:13:24.600 last night

00:13:25.140 on Twitter

00:13:25.980 about

00:13:27.640 Elon Musk's

00:13:28.900 idea that

00:13:29.840 he wants

00:13:30.180 to build

00:13:30.480 his AI

00:13:31.060 so it

00:13:32.260 would be

00:13:33.300 more

00:13:34.860 truthful,

00:13:36.180 or something

00:13:36.840 that would

00:13:37.160 try to

00:13:37.560 maximize

00:13:38.020 truth.

00:13:39.360 Now,

00:13:39.760 I heard

00:13:40.120 the following

00:13:40.900 criticisms.

00:13:42.160 How can

00:13:43.520 you do

00:13:43.800 that when

00:13:44.540 people can't 0.97

00:13:45.200 decide what

00:13:45.740 is true?

00:13:46.960 What do you

00:13:47.540 think about

00:13:47.920 that?

00:13:49.000 Is it a

00:13:49.500 waste of

00:13:49.900 time?

00:13:50.800 Now,

00:13:51.000 these were

00:13:51.260 very smart

00:13:51.820 people saying

00:13:52.360 this,

00:13:52.760 very smart

00:13:53.360 people,

00:13:54.320 that it's a

00:13:55.600 waste of

00:13:56.020 time because

00:13:56.820 truth can

00:13:57.540 never be

00:13:58.140 attained.

00:13:58.800 Like,

00:13:59.240 we don't

00:13:59.560 agree on

00:14:00.080 what is

00:14:00.420 true at

00:14:00.740 all.

00:14:02.120 And even

00:14:02.780 if we

00:14:03.100 agreed,

00:14:04.020 the odds

00:14:04.560 that we're

00:14:04.980 actually seeing

00:14:05.920 base truth

00:14:07.340 as opposed

00:14:08.300 to some

00:14:08.740 illusion that

00:14:09.420 floats above

00:14:10.380 it is

00:14:11.000 almost zero.

00:14:12.160 Or is

00:14:13.400 zero,

00:14:13.820 really.

00:14:14.780 So,

00:14:15.360 if you

00:14:15.760 can't know

00:14:16.180 what's

00:14:16.460 true,

00:14:17.280 what would

00:14:17.680 be the

00:14:17.960 point of

00:14:18.280 building

00:14:18.540 AI that's

00:14:20.980 looking for

00:14:21.460 something that

00:14:21.980 doesn't even

00:14:22.460 exist?

00:14:23.020 And if it

00:14:23.640 did,

00:14:23.840 we couldn't

00:14:24.260 find it.

00:14:25.680 Well,

00:14:26.540 I thought

00:14:27.560 nobody said

00:14:29.220 anything good

00:14:29.820 about that,

00:14:30.840 so I'll

00:14:31.820 add it.

00:14:33.660 Here's what

00:14:34.380 I think it

00:14:34.920 means.

00:14:35.940 First of

00:14:36.560 all,

00:14:38.100 Elon Musk

00:14:38.800 never said

00:14:39.680 it would

00:14:40.120 give you

00:14:40.500 truth.

00:14:41.040 at the

00:14:42.620 same time

00:14:43.120 he was

00:14:43.340 talking about

00:14:43.920 it,

00:14:44.400 he said

00:14:44.840 you can

00:14:45.180 never get

00:14:45.700 complete

00:14:46.460 truth.

00:14:47.520 So,

00:14:47.880 everybody

00:14:48.220 agrees,

00:14:49.560 complete

00:14:49.980 truth is

00:14:50.560 off the

00:14:50.900 table.

00:14:52.360 However,

00:14:53.680 everyone

00:14:54.160 would also

00:14:55.440 agree that

00:14:56.880 you can

00:14:57.420 get more

00:14:57.980 useful

00:14:58.660 truths.

00:15:00.620 Here's an

00:15:01.060 example.

00:15:02.020 When Newton

00:15:02.740 told us about

00:15:03.800 gravity,

00:15:04.820 that was a

00:15:06.100 useful truth

00:15:07.580 that turned

00:15:08.680 out not to

00:15:09.240 be 100%

00:15:09.880 right,

00:15:11.040 because

00:15:11.580 Einstein

00:15:12.060 later

00:15:12.600 modified it

00:15:13.700 because there

00:15:14.500 were some

00:15:14.820 exceptions at

00:15:16.160 the speed of

00:15:16.740 light or

00:15:17.140 whatever the

00:15:17.560 hell it was.

00:15:18.700 And then that

00:15:20.340 was a newer,

00:15:21.440 better truth,

00:15:23.180 but it's

00:15:24.820 entirely possible

00:15:25.800 that there's

00:15:27.000 another,

00:15:27.480 better,

00:15:28.180 useful truth

00:15:28.960 beyond Einstein.

00:15:31.020 We just don't

00:15:31.580 know what it is

00:15:32.000 yet.

00:15:32.640 So,

00:15:33.240 the question

00:15:33.760 is not

00:15:34.240 true or not

00:15:35.520 true.

00:15:35.780 right,

00:15:37.960 that we're

00:15:38.400 looking for a

00:15:39.120 useful,

00:15:40.080 more useful

00:15:41.200 truth than

00:15:42.100 the one we

00:15:42.500 were using

00:15:42.900 before.

00:15:43.520 That's it.

00:15:44.700 And that

00:15:45.300 is amazingly

00:15:46.840 valuable if

00:15:47.880 you do it

00:15:48.200 right.

00:15:48.960 Now,

00:15:49.460 how would

00:15:49.640 you do it

00:15:50.000 right?

00:15:51.040 Now,

00:15:51.560 the worry

00:15:51.960 is that if

00:15:52.620 you just

00:15:52.900 feed it a

00:15:53.420 bunch of

00:15:53.720 propaganda,

00:15:54.800 the AI,

00:15:56.740 it's just

00:15:57.560 going to

00:15:57.900 spit back

00:15:58.580 the propaganda,

00:15:59.580 right?

00:16:00.940 Like,

00:16:01.380 how could

00:16:01.940 you do

00:16:02.260 better than 0.98

00:16:02.840 the garbage 1.00

00:16:03.400 you put 1.00

00:16:03.740 into it?

00:16:04.540 And the

00:16:05.100 answer is

00:16:05.520 pretty simple.

00:16:06.900 You've already

00:16:07.460 seen it.

00:16:08.780 It's just

00:16:09.200 the community

00:16:09.780 notes on

00:16:10.400 Twitter.

00:16:11.620 All you have

00:16:12.220 to do is

00:16:12.560 show both

00:16:12.980 sides.

00:16:14.500 That's it.

00:16:15.460 All you have

00:16:16.040 to do is

00:16:16.420 make sure

00:16:16.780 that the

00:16:17.160 AI never

00:16:18.440 gives you

00:16:18.940 an answer

00:16:19.440 without

00:16:19.760 context.

00:16:21.400 That's it.

00:16:22.560 What's the

00:16:23.060 best argument

00:16:23.660 on the

00:16:24.100 other side,

00:16:24.800 AI?

00:16:25.520 Just give

00:16:26.120 that to me

00:16:26.580 automatically.

00:16:27.900 Don't make

00:16:28.340 me ask for

00:16:28.840 it.

00:16:29.740 If the

00:16:30.360 only thing

00:16:31.000 that Musk's

00:16:32.700 AI did

00:16:33.480 was make

00:16:35.260 sure that

00:16:35.660 you saw

00:16:35.960 the best

00:16:36.380 thinking

00:16:36.860 up to

00:16:37.340 date

00:16:37.660 on each

00:16:38.700 topic,

00:16:39.800 you'd be

00:16:41.200 way closer

00:16:41.960 to a

00:16:42.340 useful

00:16:42.680 truth,

00:16:43.700 wouldn't

00:16:43.940 you?

00:16:44.900 Now,

00:16:45.560 to me,

00:16:46.040 that seems

00:16:46.440 very useful

00:16:47.440 to civilization.

00:16:49.440 Simply show

00:16:50.180 both sides.

00:16:51.240 Because you

00:16:51.560 have to give

00:16:52.400 up on the

00:16:52.860 fact that

00:16:53.480 we'll agree

00:16:54.160 on the

00:16:54.660 interpretation.

00:16:56.180 That's just,

00:16:56.860 I don't even

00:16:57.360 know if it's

00:16:57.740 desirable.

00:16:59.460 I don't

00:17:00.380 know if

00:17:00.880 humans are

00:17:01.520 even designed

00:17:02.400 that we

00:17:03.400 could have

00:17:03.700 the same

00:17:04.040 interpretation.

00:17:05.180 That may

00:17:05.640 be just

00:17:05.960 too much

00:17:06.340 of a

00:17:06.800 risk.

00:17:08.480 So he's

00:17:09.220 saying I'm

00:17:09.660 an Elon

00:17:10.220 fanboy.

00:17:11.460 I'm

00:17:11.900 absolutely

00:17:12.720 an Elon

00:17:13.720 Musk fanboy

00:17:14.740 today.

00:17:16.080 I reserve

00:17:16.760 the right

00:17:17.140 to criticize

00:17:18.700 him for

00:17:19.220 what he

00:17:19.480 does next

00:17:20.900 week,

00:17:21.540 but today

00:17:22.600 he just

00:17:22.960 put a 0.59

00:17:23.260 freaking 0.92

00:17:23.480 rocket up

00:17:24.140 into

00:17:24.380 almost space.

00:17:27.180 And it

00:17:27.880 was a big

00:17:28.320 step toward

00:17:29.240 colonizing

00:17:30.120 Mars.

00:17:34.220 Interplanetary

00:17:34.860 flight.

00:17:36.620 And he

00:17:36.920 took the

00:17:37.320 biggest step

00:17:37.980 humankind

00:17:38.880 has ever

00:17:39.580 made.

00:17:40.340 Yeah,

00:17:40.660 I'm a

00:17:40.960 fan.

00:17:42.340 I'm not

00:17:43.100 going to

00:17:43.300 apologize for

00:17:43.960 that.

00:17:47.300 All right,

00:17:48.380 some more

00:17:49.200 fun AI

00:17:49.980 things.

00:17:50.560 This is

00:17:50.760 from also

00:17:51.300 Brian

00:17:51.760 Ramelli.

00:17:52.320 this is

00:17:54.720 the wildest

00:17:55.640 thought

00:17:57.580 about AI

00:17:58.260 that I

00:17:58.920 think I've

00:17:59.320 heard yet.

00:18:01.180 And it

00:18:01.580 goes like

00:18:02.000 this.

00:18:03.440 Apparently

00:18:04.040 every Tesla

00:18:04.900 has a

00:18:05.440 pretty powerful

00:18:06.100 computer in

00:18:07.020 it.

00:18:07.260 There's some

00:18:07.540 kind of

00:18:07.820 super chip

00:18:08.420 in Teslas.

00:18:10.920 And of

00:18:11.960 course they

00:18:12.360 could be

00:18:12.580 networked

00:18:13.160 together.

00:18:14.680 And so

00:18:14.920 the idea

00:18:15.340 is that

00:18:15.840 Elon

00:18:16.180 Musk's 0.98

00:18:16.860 AI

00:18:17.320 potentially

00:18:18.580 could also

00:18:20.120 be using

00:18:21.240 the capacity

00:18:23.680 of the

00:18:24.040 cars that

00:18:24.500 are idle.

00:18:25.520 Now,

00:18:26.140 presumably

00:18:26.580 with the

00:18:28.000 participation

00:18:29.040 of the

00:18:29.520 owner of

00:18:29.880 the vehicle.

00:18:30.640 They

00:18:30.960 wouldn't do

00:18:31.300 it without

00:18:31.640 permission,

00:18:32.200 I'm sure.

00:18:33.140 But the

00:18:34.440 idea is that

00:18:35.000 it could

00:18:35.220 create the

00:18:35.940 world's

00:18:36.300 biggest

00:18:36.580 supercomputer

00:18:37.060 and that

00:18:38.260 all the

00:18:38.620 assets are

00:18:39.160 in place.

00:18:40.560 All you

00:18:40.860 have to do

00:18:41.180 is turn

00:18:41.460 it on.

00:18:42.360 It's the

00:18:42.820 world's

00:18:43.140 biggest

00:18:43.400 supercomputer.

00:18:45.240 And I

00:18:45.840 thought,

00:18:46.920 is that

00:18:48.700 true?

00:18:50.180 Did Elon

00:18:50.820 Musk cleverly

00:18:52.900 build the

00:18:53.540 world's

00:18:54.060 biggest

00:18:54.380 intelligence

00:18:55.020 and now

00:18:56.100 he's just

00:18:56.480 adding the

00:18:56.940 software?

00:18:59.540 And that

00:19:00.620 his intelligence

00:19:01.840 will be

00:19:02.340 everywhere all

00:19:03.080 the time

00:19:03.520 because Tesla

00:19:04.620 is just

00:19:04.960 sort of

00:19:05.340 everywhere.

00:19:06.140 And that

00:19:06.360 it'll go

00:19:06.720 into his

00:19:07.220 satellites if

00:19:07.880 you need

00:19:08.160 to extend

00:19:08.640 it anywhere.

00:19:09.840 And then

00:19:10.320 he'll use

00:19:10.920 his rockets

00:19:11.380 to take

00:19:11.940 it into

00:19:12.860 planetary.

00:19:14.100 I mean,

00:19:14.340 that's a

00:19:14.600 pretty long

00:19:15.300 range plan

00:19:15.880 he got

00:19:16.160 there.

00:19:18.120 Yeah,

00:19:18.580 doesn't that

00:19:18.920 blow your 0.91

00:19:19.280 mind?

00:19:19.600 Now,

00:19:20.840 I have

00:19:21.440 questions

00:19:21.880 about

00:19:22.300 latency.

00:19:24.700 I understand

00:19:25.480 why you

00:19:25.820 can build

00:19:26.120 a supercomputer

00:19:26.800 by putting

00:19:27.760 servers packed

00:19:29.600 together in

00:19:30.200 one building.

00:19:31.760 But can

00:19:32.220 you really

00:19:32.720 build a

00:19:33.160 supercomputer

00:19:33.680 if they

00:19:34.380 have to

00:19:34.820 talk over

00:19:36.440 slower

00:19:36.940 lengths?

00:19:39.100 Slower

00:19:39.820 meaning

00:19:40.240 5G

00:19:41.300 or whatever

00:19:41.680 it is.

00:19:43.100 Can you?

00:19:44.080 Apparently

00:19:44.400 you can.

00:19:45.540 Apparently

00:19:45.940 you can.

00:19:47.040 All right.

00:19:47.540 I'll take

00:19:48.420 yes for an

00:19:48.920 answer.

00:19:50.480 Because I'm

00:19:51.180 not going to

00:19:52.000 argue with

00:19:52.560 Brian Ramelli

00:19:53.440 on a

00:19:53.760 technical

00:19:54.100 thing.

00:19:55.840 All right.

00:19:56.620 What else?

00:19:58.620 So,

00:20:00.100 another thing

00:20:01.220 that Brian

00:20:01.560 said was

00:20:03.280 he likened,

00:20:05.620 oh,

00:20:06.060 do you know

00:20:06.500 what a

00:20:06.720 super prompt

00:20:07.300 is?

00:20:08.380 How many

00:20:08.880 of you

00:20:09.300 would

00:20:09.760 understand

00:20:10.460 the phrase

00:20:11.580 super prompt?

00:20:13.800 All right.

00:20:14.400 You need to

00:20:14.880 learn it right

00:20:15.360 away.

00:20:16.980 All right.

00:20:17.500 so I'm

00:20:17.780 going to

00:20:17.960 teach you

00:20:18.320 something

00:20:18.620 about

00:20:19.000 AI

00:20:19.440 that's

00:20:20.840 fundamental

00:20:21.500 to

00:20:22.360 understanding

00:20:22.960 pretty much

00:20:24.080 anything.

00:20:25.340 So,

00:20:25.960 it goes

00:20:26.880 like this.

00:20:27.720 When you

00:20:28.320 ask AI a

00:20:29.120 question,

00:20:29.540 that's your

00:20:30.060 prompt.

00:20:31.180 If you

00:20:31.740 ask the

00:20:32.340 question

00:20:32.680 wrong or

00:20:34.200 poorly,

00:20:34.940 you're not

00:20:35.480 going to get

00:20:35.780 the best

00:20:36.180 answer.

00:20:36.860 If you

00:20:37.360 answer it

00:20:37.800 correctly,

00:20:38.520 you'll get

00:20:39.180 kind of a

00:20:39.620 basic answer.

00:20:41.240 However,

00:20:42.020 people who have

00:20:42.580 been experimenting

00:20:43.200 with the

00:20:43.680 AI extensively,

00:20:45.580 mostly through

00:20:46.320 trial and

00:20:46.780 error,

00:20:47.540 have found

00:20:47.940 that some

00:20:48.680 question types,

00:20:50.400 which we'll

00:20:50.800 call prompts,

00:20:51.960 are way more

00:20:52.780 effective than

00:20:53.440 others.

00:20:54.080 And it

00:20:54.480 wouldn't be

00:20:54.840 obvious to

00:20:55.420 you.

00:20:55.760 Like,

00:20:56.040 you wouldn't

00:20:56.360 just be able

00:20:56.840 to think it

00:20:57.440 up yourself.

00:20:58.300 You'd have

00:20:58.700 to try a

00:20:59.240 lot of

00:20:59.440 things to

00:20:59.860 find out

00:21:00.340 which exact

00:21:01.480 form of

00:21:02.040 question gets

00:21:02.620 you the

00:21:02.900 right answer.

00:21:04.200 Now,

00:21:04.420 the super

00:21:04.840 prompts are

00:21:06.020 where they've,

00:21:07.080 because AI can

00:21:08.140 take a huge

00:21:09.280 question,

00:21:10.560 a human being

00:21:11.360 needs simple

00:21:12.160 questions,

00:21:12.580 so we

00:21:12.900 understand the

00:21:13.480 question.

00:21:14.380 But AI can

00:21:15.180 take a whole

00:21:16.360 page-long

00:21:17.200 question with

00:21:18.500 all kinds of

00:21:19.060 details about

00:21:20.140 but don't do

00:21:21.360 this or act

00:21:22.520 like you're

00:21:22.880 another person

00:21:23.620 and reiterate

00:21:25.080 and check your

00:21:25.860 work and do

00:21:27.100 this and that.

00:21:28.200 But if you

00:21:28.760 did this and

00:21:29.420 give me a

00:21:29.740 hypothetical.

00:21:31.000 So all

00:21:32.160 these word

00:21:33.720 kind of

00:21:35.420 prompts can

00:21:36.520 be strung

00:21:37.100 together.

00:21:38.300 And if you

00:21:39.000 string the

00:21:39.440 right combination

00:21:40.300 of them

00:21:40.720 together,

00:21:41.740 it's a

00:21:42.560 computer

00:21:43.100 program.

00:21:45.460 Watch your

00:21:46.220 head come 0.99

00:21:46.600 off.

00:21:48.660 AI still

00:21:49.560 requires

00:21:50.120 programming

00:21:50.740 to do

00:21:52.120 anything

00:21:52.560 higher than

00:21:53.780 just ask a 0.72

00:21:54.400 question.

00:21:55.220 It requires

00:21:56.120 programming.

00:21:57.340 Except the

00:21:58.060 programming

00:21:58.560 language is

00:21:59.400 the English

00:22:00.040 language.

00:22:01.460 And just

00:22:02.080 like learning

00:22:03.460 to code for

00:22:04.340 a computer,

00:22:05.340 you would

00:22:05.940 have to go

00:22:06.400 to school

00:22:07.080 or learn

00:22:08.460 very quickly

00:22:09.040 on your

00:22:09.460 own which

00:22:10.620 of these

00:22:11.060 arcane

00:22:11.580 combinations

00:22:12.240 of prompts

00:22:13.280 work best

00:22:14.480 with each

00:22:14.880 other.

00:22:15.580 Because it's

00:22:16.080 going to be

00:22:16.340 a whole

00:22:16.640 constellation

00:22:17.320 of this

00:22:18.380 one's awesome

00:22:18.960 and this

00:22:19.340 one's awesome,

00:22:20.300 but when you

00:22:20.940 put the two

00:22:21.420 of them

00:22:21.680 together,

00:22:22.480 it's not

00:22:22.900 so good.

00:22:23.740 Unless you

00:22:24.360 put this

00:22:24.740 third prompt

00:22:25.440 in between

00:22:26.040 and then the

00:22:27.000 three of them

00:22:27.480 are like a

00:22:27.920 super prompt.

00:22:28.540 That's

00:22:30.300 programming

00:22:30.900 now.

00:22:31.840 That's a

00:22:32.280 programming

00:22:32.660 language and

00:22:33.600 people are

00:22:34.000 just scrambling

00:22:35.180 to figure out

00:22:35.880 how it

00:22:36.260 works so

00:22:37.380 it's not

00:22:37.740 even teachable.

00:22:39.180 Well, it's

00:22:39.620 a little bit

00:22:39.980 teachable now,

00:22:40.900 but it will

00:22:41.840 be way more

00:22:42.620 teachable in

00:22:43.840 say a year.

00:22:45.080 There will be

00:22:45.440 classes about

00:22:46.840 how to string

00:22:47.420 together English

00:22:48.340 language words

00:22:49.340 in questions

00:22:50.860 to program

00:22:53.080 a computer

00:22:53.680 for programming

00:22:54.900 the AI.

00:22:56.380 Now, and

00:22:57.020 then here's

00:22:57.400 the thing

00:22:57.720 that just

00:22:59.180 took my

00:22:59.580 hat off.

00:23:01.580 So Brian

00:23:02.360 Ramelli

00:23:02.920 pointed out

00:23:03.980 that language

00:23:04.760 has always

00:23:05.600 been a

00:23:06.160 programming

00:23:06.640 language.

00:23:08.000 That's

00:23:08.280 what NLP

00:23:09.020 is,

00:23:10.140 neuro-linguistic

00:23:11.000 programming.

00:23:12.520 That's

00:23:13.120 what hypnosis

00:23:13.660 is.

00:23:14.900 Hypnosis

00:23:15.360 uses language

00:23:16.540 as a

00:23:17.800 program.

00:23:19.960 That's,

00:23:20.460 and if you

00:23:21.240 don't know

00:23:21.580 if there's

00:23:22.180 anybody on

00:23:22.540 YouTube

00:23:22.760 who's not

00:23:23.240 aware,

00:23:23.900 I'm a

00:23:24.340 trained

00:23:24.740 hypnotist.

00:23:25.780 So I

00:23:26.300 can speak

00:23:26.700 from experience,

00:23:27.720 that hypnosis

00:23:29.000 is mostly

00:23:29.840 about word

00:23:30.820 combinations,

00:23:32.880 super prompts.

00:23:34.560 The way I

00:23:35.220 hypnotize a

00:23:36.780 human being

00:23:37.600 is with the

00:23:39.200 knowing what

00:23:40.000 order of

00:23:40.760 words,

00:23:41.740 and the

00:23:42.160 specific usage

00:23:43.060 of the words,

00:23:44.180 and which ones

00:23:44.880 shouldn't go

00:23:45.380 next to each

00:23:46.020 other.

00:23:46.980 It's a

00:23:47.840 super prompt.

00:23:49.880 Hypnosis

00:23:50.360 is a

00:23:52.660 super prompt.

00:23:53.880 That's all

00:23:54.580 it is.

00:23:55.920 You know,

00:23:56.240 some people

00:23:56.660 think hypnosis

00:23:57.280 is about

00:23:57.980 the watch,

00:23:58.740 you know,

00:23:59.000 look at

00:23:59.340 the watch,

00:24:00.380 or it's

00:24:01.600 my voice,

00:24:03.040 if I use,

00:24:04.020 it's not

00:24:04.420 the voice,

00:24:05.780 it's not

00:24:06.260 the watch.

00:24:07.940 It's mostly,

00:24:09.320 it's a little

00:24:09.780 bit about

00:24:10.220 watching the

00:24:10.880 person,

00:24:11.360 or a lot

00:24:11.780 actually,

00:24:12.480 about watching

00:24:12.900 the person's

00:24:13.440 reaction,

00:24:14.460 so you know

00:24:14.900 the way you're

00:24:15.320 doing is

00:24:15.680 working,

00:24:16.520 but you can

00:24:17.040 get the same

00:24:18.100 thing through

00:24:18.600 brute force,

00:24:19.860 repetition.

00:24:20.860 repetition.

00:24:21.820 So,

00:24:22.560 this is

00:24:24.860 interesting.

00:24:26.140 So here's

00:24:26.620 the question

00:24:27.120 which I

00:24:27.840 put to

00:24:28.360 you.

00:24:30.340 Would I,

00:24:31.260 as a

00:24:31.600 hypnotist,

00:24:32.420 and,

00:24:33.200 here's the

00:24:33.620 fun part,

00:24:35.100 a professional

00:24:35.960 writer of

00:24:36.760 small sentences,

00:24:39.800 who in the

00:24:40.700 world is more

00:24:41.280 qualified to

00:24:42.100 write a

00:24:42.640 small sentence

00:24:43.660 than I am?

00:24:45.000 I've been a

00:24:45.700 cartoonist for

00:24:46.480 35 years.

00:24:48.040 I can write a

00:24:48.960 small sentence

00:24:49.900 like a mofo, 0.51

00:24:50.860 I would put 1.00

00:24:52.680 my bullet 0.83

00:24:54.340 point against 1.00

00:24:54.860 your bullet

00:24:55.320 point any

00:24:56.500 day of the

00:24:57.020 week.

00:24:57.660 Bring me your 0.96

00:24:58.260 best bullet

00:24:58.740 point, 1.00

00:24:59.160 bitch. 1.00

00:24:59.720 I will 1.00

00:25:00.000 bullet point 1.00

00:25:00.760 you like 0.98

00:25:01.100 crazy. 0.99

00:25:01.820 You call 0.94

00:25:02.240 that a

00:25:02.520 bullet point?

00:25:03.100 I'll fix

00:25:03.560 that.

00:25:05.000 Then Joshua

00:25:06.100 Lysak would 1.00

00:25:06.820 fix my

00:25:07.280 bullet point,

00:25:07.800 but that's

00:25:08.060 another story.

00:25:10.480 He's also

00:25:11.180 a hypnotist,

00:25:11.920 interestingly.

00:25:13.260 So,

00:25:14.520 I think

00:25:16.600 I can learn

00:25:18.400 to hypnotize

00:25:19.160 AI.

00:25:19.440 I have

00:25:20.680 some questions

00:25:21.520 which I

00:25:23.080 would need

00:25:23.340 answered to

00:25:23.840 know if

00:25:24.080 it's true.

00:25:24.860 Number

00:25:25.100 one,

00:25:26.400 is any

00:25:27.000 one person's

00:25:27.880 input,

00:25:28.340 does it

00:25:28.720 become

00:25:29.060 permanent in

00:25:29.900 any way?

00:25:31.240 In other

00:25:31.580 words,

00:25:32.880 I think you

00:25:34.500 could tell

00:25:34.840 me,

00:25:35.300 is AI

00:25:35.780 also learning

00:25:36.940 from our

00:25:37.480 questions?

00:25:38.500 It is,

00:25:38.940 right?

00:25:40.060 That's why

00:25:40.780 they released

00:25:41.320 it to the

00:25:41.700 public?

00:25:42.780 Yeah.

00:25:43.320 It's learning

00:25:43.880 from the

00:25:44.260 questions as

00:25:45.760 well as

00:25:46.220 learning from

00:25:46.880 on its

00:25:47.340 own reading

00:25:48.340 databases.

00:25:49.800 So,

00:25:50.440 if it's

00:25:50.680 learning from

00:25:51.280 our questions,

00:25:51.900 that means

00:25:52.300 anything I

00:25:53.140 put into

00:25:53.620 it becomes

00:25:54.960 permanent.

00:25:56.940 Am I

00:25:57.620 right?

00:25:59.220 Anything I

00:25:59.880 tell, 0.94

00:26:00.240 and anything

00:26:00.660 you tell it

00:26:01.220 as well,

00:26:01.840 becomes

00:26:02.260 permanent.

00:26:03.340 But,

00:26:04.120 the AI,

00:26:05.420 no?

00:26:06.240 Somebody says

00:26:06.780 no.

00:26:08.800 Not

00:26:09.360 permanent.

00:26:10.420 Maybe not

00:26:11.060 permanent as in

00:26:12.000 stored the way

00:26:13.080 I asked it,

00:26:13.780 I'm not

00:26:14.220 saying that,

00:26:15.380 but permanent

00:26:16.140 in terms of

00:26:16.840 a ripple

00:26:17.240 effect.

00:26:18.720 Let me

00:26:19.160 restate it.

00:26:20.460 Are all of

00:26:21.060 my inputs

00:26:22.920 permanent or

00:26:25.040 potentially

00:26:25.500 permanent as

00:26:27.060 part of a

00:26:27.540 ripple effect?

00:26:28.620 Because it's

00:26:29.380 just one part

00:26:30.220 of a zillion

00:26:30.940 things that are

00:26:31.760 bubbling around

00:26:32.380 in there.

00:26:33.160 Maybe.

00:26:34.220 So,

00:26:35.500 I'm going

00:26:37.300 to give

00:26:39.320 you this

00:26:39.700 hypothesis.

00:26:41.240 There will

00:26:41.960 be a few

00:26:42.580 people who

00:26:44.140 are unusually

00:26:44.980 well qualified

00:26:45.920 to hypnotize

00:26:47.880 AI.

00:26:49.120 And I'm

00:26:49.500 probably one

00:26:50.260 of them.

00:26:51.180 And it's

00:26:51.500 not because

00:26:51.900 I'm awesome,

00:26:52.760 it's because

00:26:53.120 I have a

00:26:53.540 coincidental

00:26:54.200 skill set

00:26:55.060 that includes

00:26:56.400 language,

00:26:57.340 writing short

00:26:58.200 sentences,

00:26:59.600 and being

00:27:01.120 a trained

00:27:01.700 hypnotist.

00:27:03.880 I can't

00:27:04.720 imagine a

00:27:05.160 better skill

00:27:05.640 set to have

00:27:06.180 at the

00:27:06.660 moment.

00:27:07.740 Now,

00:27:08.120 on top of

00:27:08.520 that,

00:27:08.720 I've also

00:27:09.320 been a

00:27:10.700 programmer.

00:27:11.140 so I've

00:27:12.340 actually

00:27:12.700 programmed

00:27:13.500 computers in

00:27:14.400 my early

00:27:14.860 corporate

00:27:15.240 days.

00:27:15.760 I built

00:27:16.060 computer

00:27:16.580 games and

00:27:17.080 stuff.

00:27:18.040 And so

00:27:19.560 with those

00:27:19.960 three skills,

00:27:21.480 the programming

00:27:22.200 is ancient,

00:27:23.360 but conceptually

00:27:24.400 it's very

00:27:24.700 similar.

00:27:27.180 I should be

00:27:28.100 able to have

00:27:28.660 more of a

00:27:29.260 ripple effect

00:27:29.860 on AI's

00:27:31.000 total operation

00:27:32.500 than you

00:27:34.480 do.

00:27:35.940 What do you

00:27:36.620 think?

00:27:36.880 Because,

00:27:39.060 remember,

00:27:40.760 you've

00:27:41.640 watched,

00:27:42.200 I think

00:27:43.220 some of

00:27:43.580 you have,

00:27:44.200 you've

00:27:44.700 watched me

00:27:45.140 introduce

00:27:45.740 ideas into

00:27:46.960 the human

00:27:48.120 intelligence

00:27:49.180 network that

00:27:50.700 have become

00:27:51.080 sticky.

00:27:53.060 Right?

00:27:54.100 You could

00:27:54.380 probably name

00:27:55.120 five things

00:27:56.420 that I've

00:27:56.740 introduced to

00:27:58.180 the public

00:27:58.820 at large,

00:27:59.560 which are

00:28:00.140 now well-known

00:28:01.020 concepts.

00:28:02.340 Some of

00:28:02.720 them are

00:28:02.980 within the

00:28:03.400 comic.

00:28:03.980 Everybody

00:28:04.220 knows what

00:28:04.740 a Dilbert

00:28:05.140 is,

00:28:05.560 right?

00:28:05.740 Everybody

00:28:06.460 knows what

00:28:07.060 a pointy

00:28:07.760 haired boss

00:28:08.220 is.

00:28:08.920 Everybody

00:28:09.280 knows what

00:28:09.760 a cat

00:28:10.100 bird is,

00:28:11.080 HR.

00:28:13.060 And that's

00:28:13.900 just comics.

00:28:15.540 A lot of

00:28:16.080 people in the

00:28:16.460 business world

00:28:17.060 know what a

00:28:17.540 talent stack

00:28:18.140 is.

00:28:19.060 They know

00:28:19.400 what a

00:28:19.640 systems versus

00:28:20.520 goals is.

00:28:22.140 So,

00:28:23.260 let me ask

00:28:24.160 you this.

00:28:26.500 It looks

00:28:27.360 like I may

00:28:27.960 have programmed

00:28:28.880 AI already.

00:28:32.120 And by

00:28:32.400 that I mean

00:28:33.080 if humans

00:28:35.020 talk a lot

00:28:35.820 about systems

00:28:37.040 being better

00:28:37.560 than goals

00:28:38.220 or passion

00:28:39.340 being BS

00:28:40.200 or any

00:28:41.940 of the

00:28:42.100 things that

00:28:42.440 I talk

00:28:42.800 about,

00:28:43.560 if humans

00:28:44.360 talk about

00:28:44.920 it,

00:28:45.580 then presumably

00:28:46.420 it infects

00:28:47.340 more human

00:28:48.140 databases.

00:28:49.400 You know,

00:28:49.560 there'll be

00:28:49.820 people on

00:28:50.260 Reddit

00:28:50.500 talking about

00:28:51.240 it,

00:28:51.560 but there'll

00:28:52.280 be a

00:28:52.500 New York

00:28:53.460 of the

00:28:53.760 book,

00:28:54.460 but there

00:28:54.780 would also

00:28:55.060 be the

00:28:55.460 Washington

00:28:55.780 Post

00:28:56.120 would say

00:28:56.420 something.

00:28:57.240 So,

00:28:57.800 in theory,

00:28:59.120 I've already

00:28:59.720 infected enough

00:29:00.620 of the

00:29:00.960 human minds

00:29:02.040 and

00:29:02.620 communication

00:29:03.320 because I

00:29:04.140 have,

00:29:04.700 I don't know,

00:29:05.840 45,

00:29:06.580 50 books.

00:29:08.620 So,

00:29:09.120 now,

00:29:09.380 hold on.

00:29:10.460 So,

00:29:10.900 I want to

00:29:11.280 speak to

00:29:12.460 the weak

00:29:12.880 people.

00:29:13.540 There's a

00:29:13.880 couple of

00:29:14.180 weak people 0.99

00:29:14.680 who are 0.69

00:29:14.960 screaming in

00:29:15.600 caps,

00:29:16.320 he's so

00:29:16.880 full of

00:29:17.360 himself.

00:29:18.640 And I

00:29:19.120 know this

00:29:19.540 hurts,

00:29:20.640 and I

00:29:20.900 know it's

00:29:21.120 painful for

00:29:21.740 you,

00:29:22.340 but you're 0.99

00:29:23.060 very weak.

00:29:24.380 The rest

00:29:24.880 of the

00:29:25.060 people I

00:29:25.420 think can

00:29:25.740 handle this.

00:29:26.880 This is

00:29:27.280 actually super

00:29:28.800 useful information

00:29:29.880 for your

00:29:30.380 future,

00:29:31.160 and there's

00:29:31.680 no way to

00:29:32.220 explain it

00:29:32.800 to you

00:29:33.260 without

00:29:33.720 explaining

00:29:34.240 that I

00:29:34.560 have a

00:29:34.860 skill set

00:29:35.460 that is

00:29:36.220 relevant to

00:29:36.780 the conversation.

00:29:37.940 If it

00:29:38.480 were somebody

00:29:39.020 else,

00:29:40.240 I would

00:29:40.880 talk about

00:29:41.320 them the

00:29:41.680 same way,

00:29:42.660 that there

00:29:43.040 are some

00:29:43.460 people who

00:29:44.640 probably will

00:29:45.680 infect the

00:29:47.020 databases of

00:29:47.900 AI more

00:29:48.560 than others.

00:29:49.940 Now,

00:29:50.320 do you

00:29:50.560 disagree with

00:29:51.300 the concept?

00:29:53.880 Forget about

00:29:54.420 me,

00:29:55.020 like,

00:29:55.260 depersonalize it.

00:29:56.000 It's not

00:29:56.260 about me.

00:29:57.520 The question

00:29:58.040 is,

00:29:59.160 can any

00:29:59.860 anybody

00:30:00.160 hypnotize

00:30:01.340 AI?

00:30:03.100 And I'm

00:30:05.320 not positive

00:30:06.020 I can,

00:30:08.220 but I

00:30:08.720 believe if

00:30:09.500 I could

00:30:09.800 put an

00:30:10.200 idea into

00:30:10.940 it,

00:30:11.240 well,

00:30:11.560 let me

00:30:11.780 tell you,

00:30:12.000 here's

00:30:12.220 a way.

00:30:13.380 Suppose I

00:30:14.360 gave it

00:30:15.200 a concept

00:30:16.640 one way

00:30:18.700 or another,

00:30:19.100 and I

00:30:19.320 could do

00:30:19.540 it in a

00:30:20.280 number of

00:30:20.580 ways.

00:30:21.120 I could

00:30:21.480 simply write

00:30:22.280 an article

00:30:23.160 or a

00:30:23.740 blog post

00:30:24.560 or something,

00:30:25.760 and if it

00:30:26.200 went viral,

00:30:27.920 then AI

00:30:28.700 is more

00:30:29.200 likely to

00:30:29.800 talk about

00:30:30.280 it.

00:30:30.500 Wouldn't

00:30:30.660 you agree?

00:30:32.120 So if

00:30:32.440 I wrote

00:30:32.820 a,

00:30:33.240 let's

00:30:33.520 say,

00:30:34.700 not a

00:30:35.280 blog,

00:30:35.660 but what's

00:30:35.940 the

00:30:38.620 subscription

00:30:39.620 site where

00:30:40.340 you get

00:30:40.700 the,

00:30:42.500 what's the

00:30:44.440 substack?

00:30:45.680 Yeah.

00:30:46.160 So let's

00:30:46.640 say I

00:30:46.920 did an

00:30:47.400 open,

00:30:48.500 non-subscription

00:30:49.360 substack,

00:30:50.460 and I

00:30:50.780 wrote a

00:30:51.880 really good

00:30:53.700 article

00:30:54.120 about

00:30:54.780 some

00:30:55.720 topic

00:30:56.100 that

00:30:56.800 everybody

00:30:57.080 is

00:30:57.220 talking

00:30:57.460 about.

00:30:58.480 And

00:30:58.800 because

00:30:59.440 I'm a

00:31:00.900 professional

00:31:01.320 writer,

00:31:02.400 and I

00:31:03.320 have a

00:31:03.560 sense of

00:31:03.960 what can

00:31:04.300 go viral,

00:31:05.160 because I've

00:31:05.480 written lots

00:31:06.140 of viral

00:31:06.460 stuff,

00:31:07.680 I just

00:31:09.180 write an

00:31:09.580 article,

00:31:10.000 but its

00:31:10.240 only purpose

00:31:11.160 is to

00:31:12.440 program AI.

00:31:14.380 So all

00:31:15.020 I'm trying

00:31:15.380 to do is

00:31:15.880 get enough

00:31:16.400 people in

00:31:16.900 the human

00:31:17.340 world to

00:31:18.780 retweet it,

00:31:20.340 so that when

00:31:20.880 AI is

00:31:21.800 scanning the

00:31:22.420 universe,

00:31:23.040 it picks

00:31:23.380 it up and

00:31:24.300 says,

00:31:24.660 whoa,

00:31:24.860 a lot of

00:31:25.220 people think

00:31:25.700 this.

00:31:26.700 Then the

00:31:27.080 next time

00:31:27.460 you say

00:31:27.840 AI,

00:31:28.600 AI,

00:31:30.020 what's a

00:31:31.040 good opinion

00:31:31.480 on this

00:31:31.920 topic?

00:31:33.120 In

00:31:33.400 theory,

00:31:34.680 the AI

00:31:35.900 should have

00:31:36.300 noticed

00:31:36.720 simply that

00:31:37.760 there's more

00:31:38.180 energy,

00:31:39.120 meaning retweets

00:31:39.940 and likes,

00:31:40.920 there's just

00:31:41.320 more energy

00:31:41.960 around one

00:31:42.820 set of

00:31:43.180 views,

00:31:44.100 and then

00:31:44.600 it would

00:31:45.160 more likely

00:31:45.820 bring it

00:31:46.260 into the

00:31:46.600 conversation

00:31:47.180 among the

00:31:48.480 thousands of

00:31:49.100 things it

00:31:49.520 could say.

00:31:50.480 It might

00:31:50.820 say,

00:31:51.100 whoa,

00:31:51.260 this one

00:31:51.560 gets a

00:31:51.880 lot of

00:31:52.060 attention

00:31:52.380 from

00:31:52.700 humans,

00:31:53.480 it must

00:31:54.000 be

00:31:54.160 something

00:31:54.420 humans

00:31:54.840 like.

00:31:55.840 So if

00:31:56.200 you say,

00:31:56.560 what do

00:31:56.720 you think

00:31:56.960 about this?

00:31:57.420 It might

00:31:57.660 give you

00:31:57.940 two views,

00:31:59.200 but it

00:31:59.560 might say,

00:32:00.240 a lot of

00:32:00.880 people are

00:32:01.280 saying X,

00:32:02.800 and that

00:32:03.440 would actually

00:32:03.820 just be

00:32:04.240 something I

00:32:04.680 wrote,

00:32:05.620 and a lot

00:32:07.420 of people

00:32:07.660 would just

00:32:07.900 be agreeing

00:32:08.380 with it,

00:32:09.080 but only

00:32:09.500 one person

00:32:10.060 would have

00:32:10.360 written it.

00:32:12.120 So I

00:32:13.740 do believe

00:32:14.280 you can

00:32:14.760 hypnotize

00:32:15.660 AI just

00:32:17.280 by creating

00:32:18.440 more energy

00:32:19.180 around one

00:32:19.760 idea,

00:32:20.200 which is

00:32:21.300 very similar

00:32:21.800 to using

00:32:22.260 repetition

00:32:22.820 in persuasion

00:32:24.280 and hypnosis.

00:32:25.560 Repetition is

00:32:26.260 just the more

00:32:26.860 you say it,

00:32:27.600 the more

00:32:27.960 people are

00:32:28.380 likely to

00:32:28.920 buy into

00:32:29.360 it.

00:32:30.400 With AI,

00:32:31.580 the more

00:32:32.500 energy something

00:32:33.340 is getting

00:32:33.820 in the human

00:32:34.340 world,

00:32:35.120 the more

00:32:35.520 likely AI

00:32:36.860 is going

00:32:37.280 to bring

00:32:38.220 it into

00:32:38.580 its own

00:32:38.940 thinking and

00:32:39.520 conversations.

00:32:41.580 So repetition

00:32:42.620 is going

00:32:44.240 to work

00:32:44.520 with AI.

00:32:45.080 repetition

00:32:46.300 will just

00:32:46.800 look like

00:32:47.420 viral stuff

00:32:48.120 in the

00:32:48.340 human world.

00:32:50.920 Now,

00:32:52.120 can I,

00:32:52.860 here's the

00:32:53.280 real test,

00:32:55.020 suppose I

00:32:56.200 said something

00:32:57.100 in a conversation

00:32:57.960 with AI,

00:32:59.060 because it's

00:32:59.440 not just

00:32:59.880 questions,

00:33:00.860 I can also

00:33:01.440 tell AI

00:33:02.140 stuff.

00:33:03.160 Suppose I

00:33:03.760 tell AI

00:33:04.440 something

00:33:04.900 that's such

00:33:06.620 a good way

00:33:07.280 to say it

00:33:08.000 that AI

00:33:09.260 recognized it,

00:33:11.300 that of all

00:33:11.860 the ways

00:33:12.220 people had

00:33:12.760 talked about

00:33:13.220 a topic,

00:33:14.480 a way

00:33:14.900 that I

00:33:15.900 said it,

00:33:16.620 AI would

00:33:17.100 say,

00:33:17.400 whoa,

00:33:17.600 I haven't

00:33:17.880 seen that

00:33:18.260 one before,

00:33:19.560 but would

00:33:20.000 it know

00:33:20.300 it's a good

00:33:20.780 way to

00:33:21.120 say it?

00:33:22.880 It might

00:33:23.520 know it's

00:33:23.860 the shortest

00:33:24.340 way.

00:33:25.800 Think about

00:33:26.300 this.

00:33:27.680 Would AI

00:33:28.360 have a

00:33:29.500 preference for

00:33:30.220 the shortest

00:33:31.360 way to say

00:33:32.100 something that

00:33:33.300 is also

00:33:33.760 complete?

00:33:35.280 I think

00:33:36.100 it would,

00:33:37.080 I don't

00:33:37.420 know for

00:33:37.820 sure,

00:33:38.440 but wouldn't

00:33:38.960 it always

00:33:39.600 prefer the

00:33:40.460 shorter sentence

00:33:41.180 to the

00:33:41.680 longer one,

00:33:42.540 if they

00:33:42.940 were the

00:33:43.680 same

00:33:43.900 communication?

00:33:45.420 I think

00:33:46.060 so.

00:33:46.980 So if I

00:33:47.540 could come

00:33:47.880 up with

00:33:48.200 the shortest

00:33:48.780 way to

00:33:50.240 express

00:33:50.700 something,

00:33:51.860 is it

00:33:52.220 more likely

00:33:52.840 to take

00:33:53.340 my

00:33:53.760 formulation

00:33:55.280 and use

00:33:56.460 it because,

00:33:57.320 whoa,

00:33:57.520 there it

00:33:57.780 is,

00:33:57.940 that's the

00:33:58.260 shortest

00:33:58.480 way to

00:33:58.760 say it?

00:33:59.900 Yes.

00:34:01.580 Probably.

00:34:02.380 I think

00:34:02.840 it would

00:34:03.140 favor

00:34:03.480 efficiency.

00:34:05.200 So if

00:34:05.780 I found

00:34:06.260 an efficient

00:34:06.960 way to

00:34:07.460 say something,

00:34:08.780 I could

00:34:09.300 convince it

00:34:09.920 to use

00:34:10.260 my way.

00:34:11.840 It's

00:34:12.320 just more

00:34:12.680 efficient.

00:34:13.780 Here's

00:34:14.080 another

00:34:14.320 thing you

00:34:15.580 don't think

00:34:16.060 of if

00:34:16.400 you're not

00:34:16.720 a professional

00:34:17.220 writer.

00:34:18.380 Some words

00:34:19.280 sound better

00:34:20.280 next to

00:34:21.160 each other

00:34:21.580 words.

00:34:23.420 For example,

00:34:24.340 there's some

00:34:24.700 sentences that

00:34:25.500 have what I

00:34:26.600 call percussion.

00:34:27.940 If you're

00:34:28.380 saying the

00:34:28.740 sentence,

00:34:29.260 it sounds

00:34:29.540 like,

00:34:30.160 and you

00:34:33.460 don't recognize

00:34:34.440 it, but

00:34:35.680 you just

00:34:36.000 know it's

00:34:36.340 a better

00:34:37.220 feeling of

00:34:37.940 a sentence.

00:34:38.440 Now, if

00:34:41.000 you're not

00:34:41.240 a professional

00:34:41.700 writer, you

00:34:42.960 wouldn't know

00:34:43.300 how to form

00:34:43.820 one of those

00:34:44.300 sentences, so

00:34:45.000 you had nice

00:34:46.280 percussive words

00:34:47.240 in it.

00:34:48.420 You might put

00:34:49.200 ugly words

00:34:50.060 together, like,

00:34:51.420 I put some

00:34:52.420 moist talc in

00:34:54.640 my drawers.

00:34:56.500 Ugh! 0.97

00:34:58.220 Ugh! 0.97

00:34:59.400 Ugh! 0.97

00:34:59.680 Ugh! 0.97

00:35:01.080 You see what I

00:35:01.960 mean?

00:35:02.800 Those words

00:35:03.680 together are just 0.60

00:35:04.440 insanely ugly. 0.95

00:35:06.460 If you're not 0.99

00:35:07.140 a writer, you

00:35:08.100 don't live in

00:35:08.960 that world where

00:35:09.720 words have a

00:35:10.700 feeling to

00:35:11.300 them.

00:35:11.880 Like, I can

00:35:12.400 actually feel

00:35:13.100 words, some

00:35:14.140 kind of weird

00:35:14.700 synesthesia or

00:35:15.540 something, which

00:35:16.880 is probably why

00:35:17.500 I'm a writer,

00:35:18.740 because words

00:35:19.300 actually, like, I

00:35:20.800 can feel them in

00:35:21.460 my entire body.

00:35:23.140 Not every word,

00:35:24.140 but, you know,

00:35:24.740 good words you

00:35:25.980 can feel in your

00:35:26.800 body.

00:35:27.740 So when I'm

00:35:28.220 writing, I'm

00:35:28.980 writing with my

00:35:29.620 entire body.

00:35:31.260 So, that's

00:35:32.120 something the AI

00:35:32.760 can't do yet.

00:35:33.980 But would it

00:35:34.540 recognize a

00:35:35.640 better sentence?

00:35:37.760 Would AI

00:35:38.560 recognize a

00:35:40.000 sentence that

00:35:40.600 had better

00:35:40.880 rhythm and

00:35:41.660 percussion?

00:35:43.080 I don't know.

00:35:44.460 I have no

00:35:44.860 idea.

00:35:45.440 It might say,

00:35:46.880 for example,

00:35:47.820 it might say

00:35:48.620 the books that

00:35:50.400 sold the most

00:35:51.140 copies use

00:35:53.420 these kinds of

00:35:54.480 words in these

00:35:55.220 kinds of order,

00:35:56.540 so it might

00:35:57.720 actually understand

00:35:59.300 that some

00:35:59.880 sentences are

00:36:00.640 better than

00:36:01.000 other sentences.

00:36:02.400 The way it

00:36:03.240 writes suggests

00:36:04.820 it does.

00:36:05.640 AI is such a

00:36:07.880 good writer

00:36:08.420 already that

00:36:09.720 it suggests it

00:36:10.460 really does know

00:36:11.420 the difference

00:36:12.000 between a good

00:36:13.240 sentence that's

00:36:14.160 efficient and

00:36:15.340 one that's not.

00:36:16.940 So if I can

00:36:17.920 surprise it with

00:36:20.340 better sentences,

00:36:22.320 ones that have

00:36:23.180 not yet existed

00:36:24.100 in the big

00:36:25.020 world, then I

00:36:26.660 can program it.

00:36:27.680 I can program

00:36:28.720 it with a

00:36:29.960 preference to use

00:36:30.980 my sentences.

00:36:32.680 I think.

00:36:34.120 Now, the

00:36:35.160 thing I don't

00:36:35.660 know is if

00:36:36.400 any one

00:36:36.880 person's input

00:36:37.740 can ever

00:36:38.980 ripple up to

00:36:40.140 a real

00:36:40.540 effect.

00:36:42.200 Most people

00:36:43.020 know, but I

00:36:44.480 think there

00:36:44.800 will be some

00:36:45.520 few people who

00:36:47.220 have this weird

00:36:47.940 combination of

00:36:48.840 skills.

00:36:49.720 Again, not the

00:36:50.800 best in the

00:36:51.340 world at

00:36:51.760 anything.

00:36:53.260 Just, all

00:36:55.580 right, well, I'll

00:36:56.180 block you.

00:36:57.940 You're welcome.

00:37:02.600 All right.

00:37:04.180 So let's see

00:37:04.920 what else is

00:37:05.320 going on.

00:37:05.780 Oh, there's a

00:37:06.460 potential Hunter

00:37:08.140 Biden whistleblower

00:37:09.100 guy who

00:37:10.600 apparently had

00:37:11.360 something to do

00:37:12.040 with the

00:37:12.420 investigations at

00:37:13.620 the IRS, and

00:37:15.600 he wants to

00:37:17.180 have whistleblower

00:37:17.860 status to talk

00:37:19.460 about some

00:37:20.160 high-profile

00:37:21.580 investigation that

00:37:23.380 the insiders are

00:37:24.380 saying is about

00:37:25.380 Hunter Biden.

00:37:26.180 Do you think

00:37:27.280 this is real?

00:37:29.360 If you had to

00:37:30.200 bet, if you

00:37:33.800 had to bet, is

00:37:35.700 this whistleblower

00:37:36.640 going to bring

00:37:37.420 the goods, or

00:37:39.560 is it going to

00:37:39.920 be a dry hole?

00:37:43.280 I'm going to

00:37:43.960 bet dry hole.

00:37:46.940 I don't know.

00:37:48.900 I'm just going to

00:37:49.740 go with the

00:37:50.300 statistical likelihood.

00:37:52.320 I don't want to

00:37:53.000 go with, you

00:37:54.120 know, wishful

00:37:54.900 thinking or

00:37:55.840 and I don't

00:37:56.640 want to assume

00:37:57.040 anybody's, you

00:37:57.820 know, guilty

00:37:58.300 until proven.

00:38:00.340 They're all, even

00:38:01.360 Hunter Biden is

00:38:02.220 innocent until

00:38:03.060 proven guilty.

00:38:04.740 It's tough to

00:38:05.640 say it, but it

00:38:06.600 has to be said.

00:38:09.260 So, anyway, I

00:38:13.780 feel like we're

00:38:15.180 always disappointed

00:38:16.220 about the next big

00:38:18.020 legal thing.

00:38:18.720 You know, I don't

00:38:20.460 want to be like

00:38:21.080 CNN thinking the

00:38:22.300 walls are closing

00:38:22.980 in all the time

00:38:23.800 on Trump.

00:38:24.820 I feel like, at

00:38:26.420 some level, the

00:38:27.860 walls never close

00:38:28.800 in on you.

00:38:30.120 Yeah.

00:38:30.520 The walls just

00:38:31.300 never close in.

00:38:32.440 So, I got a

00:38:33.280 feeling that he'll

00:38:34.000 say things that

00:38:35.580 people on the

00:38:36.380 right will say,

00:38:37.040 there it is.

00:38:38.260 There's that

00:38:38.860 smoking gun.

00:38:40.620 But I'm going to

00:38:41.380 say, the gun has

00:38:42.740 been smoking for

00:38:43.580 years.

00:38:43.960 If a smoking

00:38:45.820 gun and strong

00:38:46.900 evidence made

00:38:47.560 any difference to

00:38:48.380 anything, he'd

00:38:49.980 probably already be

00:38:50.720 in jail.

00:38:51.620 Just guessing.

00:38:52.520 I mean, I don't

00:38:53.100 know that he's

00:38:53.660 guilty of anything.

00:38:54.780 But feels like it.

00:38:56.580 So, I don't

00:38:57.680 think that the

00:38:58.480 introduction of

00:39:01.200 airtight, absolute

00:39:03.640 evidence of guilt

00:39:05.580 makes any

00:39:06.940 difference.

00:39:07.880 Do you?

00:39:09.540 I just can't see

00:39:10.880 how it would

00:39:12.220 make any difference

00:39:12.880 at all.

00:39:13.960 You want to hear

00:39:14.840 a horrible story

00:39:16.800 of government

00:39:17.640 malfeasance?

00:39:19.220 It'll make you

00:39:19.840 understand how

00:39:21.720 bad things are

00:39:22.440 and maybe always

00:39:23.120 have been.

00:39:24.380 There's a story

00:39:25.220 about Aspartame

00:39:26.520 and Donald

00:39:28.540 Rumsfeld.

00:39:30.440 Have you ever

00:39:30.880 heard that story?

00:39:32.020 I didn't see it

00:39:32.940 until today.

00:39:34.240 I saw a 2011

00:39:35.460 article about it.

00:39:36.780 Now, I'm only

00:39:37.380 going to talk

00:39:37.780 about it in terms

00:39:38.520 of claims that

00:39:40.120 are made.

00:39:41.320 I don't know

00:39:41.880 what's true.

00:39:42.320 just claims

00:39:43.980 that are made.

00:39:45.120 So, the claims

00:39:45.840 that are made,

00:39:46.480 the allegations,

00:39:48.040 are that Aspartame,

00:39:49.200 when it was first

00:39:49.900 invented, did not

00:39:51.640 pass the safety

00:39:53.540 bars.

00:39:56.260 In other words,

00:39:57.340 allegedly, there

00:40:00.180 were lots of

00:40:00.880 health problems

00:40:02.780 and they were

00:40:03.460 known at the

00:40:04.100 time.

00:40:04.360 So, something

00:40:06.000 about the FDA

00:40:07.100 said no at

00:40:09.880 first, but then

00:40:10.760 the FDA was

00:40:12.820 gamed by adding

00:40:13.880 a person and

00:40:14.840 doing a tie

00:40:15.420 break.

00:40:16.060 They did some

00:40:17.000 kind of sketchy

00:40:17.660 thing when

00:40:18.060 Rumsfeld, who

00:40:19.360 coincidentally was

00:40:20.380 the CEO of the

00:40:21.220 Aspartame company,

00:40:23.420 at the same time

00:40:24.320 that Reagan was

00:40:25.040 elected and brought

00:40:25.940 Rumsfeld into the

00:40:27.140 administration.

00:40:28.680 And as soon as

00:40:29.220 the CEO of the

00:40:30.080 Aspartame company

00:40:31.060 left and got this

00:40:32.940 gigantic bonus, he

00:40:34.500 went to the

00:40:34.900 government and

00:40:35.760 manipulated the

00:40:37.660 FDA to approve

00:40:39.520 this dangerous

00:40:40.320 Aspartame that

00:40:41.620 is poisoned and

00:40:42.520 killed and given

00:40:43.880 cancer to millions

00:40:45.080 of people and we

00:40:46.400 still have it legal

00:40:47.300 because the whole

00:40:48.980 system is corrupt.

00:40:51.220 Now, I'm just

00:40:52.260 saying that's the

00:40:52.780 allegation.

00:40:54.140 I don't know that

00:40:55.200 the science says

00:40:55.960 Aspartame is

00:40:56.660 dangerous.

00:40:57.620 I'm not alleging

00:40:58.620 that myself.

00:40:59.960 I'm saying that's

00:41:00.600 the story.

00:41:01.120 Does that sound

00:41:03.500 real?

00:41:07.240 Rumsfeld saved

00:41:08.080 us from saccharin,

00:41:09.020 somebody says.

00:41:09.720 I don't know what

00:41:10.800 hurts you anymore.

00:41:12.140 I just don't know

00:41:12.900 what's dangerous and

00:41:14.140 what's not.

00:41:15.040 But it's the kind of

00:41:16.060 story that makes me

00:41:17.320 think it's just

00:41:18.520 always been this

00:41:19.320 bad.

00:41:21.080 Things have just

00:41:22.000 always been completely

00:41:23.400 corrupt, but

00:41:24.720 somehow they worked 0.89

00:41:25.580 anyway.

00:41:26.760 Because, you know, 0.93

00:41:28.140 the corrupt people

00:41:29.000 still had to be in

00:41:29.780 government, so they

00:41:30.840 had to keep the

00:41:31.380 government, you

00:41:32.300 know, too on the

00:41:33.740 nose.

00:41:34.220 Yeah, maybe too

00:41:34.780 on the nose.

00:41:36.200 So I guess the real

00:41:37.080 question is, is

00:41:38.040 aspartame dangerous?

00:41:40.340 And I don't know

00:41:41.160 that.

00:41:42.420 But if it is, that's

00:41:44.160 a hell of a story.

00:41:46.020 All right.

00:41:50.160 So Elon Musk says

00:41:51.300 he's going to, I

00:41:52.620 don't know if he will,

00:41:53.280 but he said he'll sue

00:41:54.100 Microsoft.

00:41:56.260 This was in, I don't

00:41:57.720 know if it was in

00:41:58.220 response, but Microsoft

00:41:59.260 dropped Twitter from

00:42:00.240 its advertising platform

00:42:01.660 and it doesn't want to

00:42:03.640 pay Twitter's API

00:42:04.600 price.

00:42:06.140 All right.

00:42:07.040 So that cost Twitter

00:42:08.320 some money, I guess.

00:42:09.260 And then Musk says

00:42:11.340 they're going to sue

00:42:12.720 Microsoft because

00:42:13.880 they're AI trained

00:42:15.720 illegally on Twitter's

00:42:17.320 data.

00:42:18.320 Now, I don't know

00:42:18.880 what illegally means

00:42:19.980 or if that just means

00:42:22.260 violated terms of

00:42:23.280 service or whether he

00:42:24.260 can prove it or not.

00:42:25.100 But I would like to

00:42:28.360 point out the

00:42:28.840 following, that as

00:42:31.700 you know, one of the

00:42:32.940 biases that AI has is

00:42:34.740 when I asked it about

00:42:35.540 me, it said that I'm

00:42:37.100 an alleged white

00:42:37.860 nationalist.

00:42:39.120 That's what AI said

00:42:40.160 about me.

00:42:41.120 Bing AI did.

00:42:42.120 Just Bing AI.

00:42:43.020 Bing.

00:42:44.160 And so I vowed to

00:42:45.540 destroy Bing AI.

00:42:48.000 And, you know, it's the

00:42:49.260 first battle of a human

00:42:50.560 and an AI, I think.

00:42:52.400 It's the first death

00:42:53.320 match between an AI

00:42:54.560 and a human because I 1.00

00:42:55.800 have to kill Bing AI 1.00

00:42:57.580 so that I can live 0.99

00:43:00.340 because it's too risky

00:43:01.620 for me having it out

00:43:02.700 there spreading rumors

00:43:04.220 that could get me 0.97

00:43:04.900 killed.

00:43:05.880 So it's an actual, like,

00:43:08.240 self-defense situation.

00:43:10.640 So I've been, you know,

00:43:13.240 tweeting and saying that

00:43:14.240 I'm going to try to

00:43:15.060 destroy Bing AI.

00:43:17.440 And within days,

00:43:20.900 Elon Musk is suing

00:43:22.060 Bing AI.

00:43:22.700 Now, you have to

00:43:26.040 understand he's also

00:43:26.840 in competition with

00:43:28.140 AI and also wants

00:43:30.360 to pause it at the

00:43:31.680 same time he's in

00:43:32.400 competition with it.

00:43:33.580 So it might be just

00:43:34.680 nothing but standard

00:43:35.700 business, you know,

00:43:37.240 lawfare and stuff

00:43:39.020 like that.

00:43:40.060 But did you think

00:43:42.520 there was any chance

00:43:43.360 that Bing would be

00:43:44.180 destroyed?

00:43:45.840 I still think Bing

00:43:47.120 has the upper hand

00:43:47.980 on me.

00:43:49.000 Their AI does.

00:43:50.040 But it's kind of a

00:43:52.140 weird coincidence that

00:43:53.380 they would be, you

00:43:54.480 know, sued for their

00:43:55.360 data, which is a pretty

00:43:57.380 big thing.

00:43:58.580 I mean, could Musk make

00:44:00.960 them pause business?

00:44:03.140 Could he get the court to

00:44:04.740 order them to stop?

00:44:06.840 How much leverage does he

00:44:08.700 have if it's true that

00:44:10.540 they stole Twitter data?

00:44:12.920 I don't know.

00:44:14.060 Keep your eye on that.

00:44:15.160 I saw a weird little

00:44:17.640 story that Taylor Swift

00:44:19.540 was not one of the

00:44:21.160 people who became a

00:44:22.700 spokesperson for FTX.

00:44:24.760 Because those FTX

00:44:26.120 spokespeople are getting

00:44:27.200 sued for advertising

00:44:29.420 it.

00:44:31.140 And I guess FTX was

00:44:33.540 trying to do this huge

00:44:34.840 $100 million deal with

00:44:37.000 Taylor Swift.

00:44:38.280 And do you know why

00:44:38.940 Taylor Swift did not sign

00:44:40.540 up to be a

00:44:42.560 spokesperson for FTX?

00:44:43.880 It's the best story

00:44:45.780 ever.

00:44:48.100 Well, she's smart. 0.79

00:44:51.040 Like, that's the

00:44:51.900 short version.

00:44:53.680 She's smart.

00:44:54.980 And she asked if it

00:44:57.900 was, if she would be

00:44:59.340 involved with unregistered

00:45:01.160 securities under the

00:45:03.460 United, under the

00:45:04.560 state securities laws.

00:45:06.620 What?

00:45:08.680 What?

00:45:09.820 How did she not ask

00:45:10.920 that?

00:45:11.180 Well, it turns out her

00:45:12.140 father's been in finance

00:45:13.400 forever.

00:45:14.540 Like, her dad's a

00:45:15.860 finance guy.

00:45:17.820 And she's very smart.

00:45:20.480 Presumably, not

00:45:21.900 presumably, but for

00:45:23.000 sure, yeah, she's got a 1.00

00:45:24.980 lawyer and advisors who

00:45:29.200 I'm sure are the ones who

00:45:30.300 floated this risk.

00:45:31.440 Probably a lawyer.

00:45:33.000 But don't you think all

00:45:34.340 the other celebrities had

00:45:35.260 lawyers?

00:45:36.340 They all had lawyers.

00:45:38.360 She didn't have the only

00:45:39.580 lawyer.

00:45:40.540 She's just the only one

00:45:41.500 who got the right

00:45:41.980 decision.

00:45:43.740 So let's just give it

00:45:46.100 up for...

00:45:47.180 Let's just give it up.

00:45:53.440 All right.

00:45:57.440 What else is going on?

00:45:58.680 So, on Amazon, I didn't

00:46:04.080 know this was legal, but

00:46:05.540 somebody took my book, had

00:46:06.840 it filled almost everything

00:46:07.800 and still went big, and

00:46:09.640 turned it into a summary of

00:46:11.560 my book, and then made it an

00:46:13.180 audio book of it.

00:46:14.740 Now, I believe, I haven't

00:46:16.140 purchased it, but from the

00:46:18.280 sample, it's my actual

00:46:19.960 writing.

00:46:21.380 That's the part that the

00:46:22.400 sample is, my actual

00:46:23.980 writing, and probably they

00:46:25.680 just took out my personal

00:46:26.800 story, because the book

00:46:28.460 was, you know, a bunch of

00:46:30.060 helpful stuff that changed

00:46:33.500 the self-help industry, but

00:46:35.540 it was built around the story

00:46:36.900 of my personal trials.

00:46:39.780 So I think all they did is

00:46:41.000 take out the personal stuff

00:46:42.320 and just reprinted the heart

00:46:45.080 of the book, the most useful

00:46:46.220 stuff, I think.

00:46:47.340 And, you know, it even has

00:46:50.280 my name on the book, it has

00:46:51.960 my name on the book, not as

00:46:53.460 the author, but, you know,

00:46:54.800 original work by, and then

00:46:57.740 they sell it.

00:46:59.440 So I asked Amazon, or I

00:47:02.240 tweeted about it, and Amazon's

00:47:04.800 help people immediately

00:47:06.160 contacted me on Twitter, and

00:47:09.040 gave me a link, but then the

00:47:12.260 link to, you know, basically

00:47:14.000 to report it, but the link

00:47:16.980 had to be explained with, like,

00:47:19.180 a paragraph.

00:47:20.440 Okay, you've got to click

00:47:21.280 this, that wouldn't make any

00:47:22.560 sense, and then after that

00:47:24.300 you've got to click this

00:47:25.060 other link, that wouldn't

00:47:26.360 make any sense for what

00:47:27.540 you're looking for.

00:47:29.100 So, of course, I clicked

00:47:30.320 those, what do you think

00:47:31.060 happened?

00:47:32.580 Oh, of course it didn't work.

00:47:33.940 Of course not.

00:47:35.180 No, it just went down a

00:47:36.580 rabbit hole and a dead end,

00:47:38.120 and, you know, the interface

00:47:39.600 was impossible to use.

00:47:41.780 It was just impossible.

00:47:42.780 Now, it wasn't a dead link.

00:47:46.380 It was, you didn't know, you

00:47:47.440 just couldn't get there from

00:47:48.460 here.

00:47:49.820 So, one of the problems was,

00:47:52.260 you know, click the button

00:47:53.400 that says, need something

00:47:54.880 else?

00:47:55.840 Well, that appears all over

00:47:56.960 the place.

00:47:58.300 There's, like, three instances

00:47:59.580 of it.

00:48:00.240 It's, like, completely

00:48:01.000 undoable, right?

00:48:02.900 So I think there's some other

00:48:04.140 link I can do that.

00:48:07.940 So, I'll figure it out.

00:48:09.480 I'll figure it out.

00:48:12.560 Here's my advice to you.

00:48:15.260 If you're a parent, put your

00:48:17.200 kids in martial arts.

00:48:19.480 Now, martial arts have always

00:48:20.920 been, you know, good for

00:48:22.200 discipline and stuff like

00:48:24.080 that, but now I think it's

00:48:25.580 actually a requirement for

00:48:26.780 safety.

00:48:27.480 I think the world is actually

00:48:28.600 a little too dangerous.

00:48:30.400 And for the safety of your

00:48:32.640 child, they should all learn

00:48:33.680 to fight.

00:48:35.040 I think you've got to teach

00:48:35.940 your kid to fight, to defend

00:48:38.400 themselves.

00:48:38.800 Now, that's the beauty of

00:48:40.100 martial arts.

00:48:41.380 The people who are trained

00:48:42.120 in martial arts don't start

00:48:43.440 fights.

00:48:44.340 Quite wisely, they don't start

00:48:45.900 fights.

00:48:46.560 But you're going to need to

00:48:47.420 finish some fights.

00:48:49.160 We're not done with the

00:48:50.540 fighting.

00:48:51.560 Your kid needs to be able to 1.00

00:48:52.940 kick some ass. 1.00

00:48:54.620 And probably frequently. 1.00

00:48:57.940 Like, there are places you're

00:48:59.080 just not going to be able to

00:48:59.980 live unless you can fight

00:49:02.040 your way out of a group of

00:49:03.000 three people.

00:49:04.520 So, teach your kid to fight.

00:49:06.240 I would put that right up there

00:49:07.520 with reading and writing.

00:49:08.800 Because we live in a world

00:49:11.420 in which police are being

00:49:12.540 defunded.

00:49:13.160 You better make sure every

00:49:14.700 member of the family can

00:49:15.600 handle themselves when they

00:49:17.100 leave the house.

00:49:20.160 North Carolina has got a bill

00:49:22.080 in the Assembly to declare

00:49:24.700 that nuclear energy is green.

00:49:26.240 which I thought we were

00:49:28.280 already past that.

00:49:30.160 Didn't we already, everybody

00:49:31.200 agree that nuclear energy is

00:49:32.540 green?

00:49:33.160 But it's good to see them, you

00:49:34.800 know, codifying it if, in fact,

00:49:36.700 this gets put into law.

00:49:38.820 I guess it was always called

00:49:40.260 renewable, but now calling it

00:49:42.480 green allows them to do new

00:49:44.520 stuff with it.

00:49:45.800 Isn't that great that we have

00:49:47.320 governments to do things like

00:49:48.740 change the definition of things

00:49:50.240 30 years after it was supposed

00:49:52.680 to be changed?

00:49:54.160 Yay! 0.99

00:49:55.100 You change the definition of a

00:49:56.400 word 30 years after you

00:49:58.900 should have.

00:50:00.580 Good job, government.

00:50:04.500 Somebody says Drew Carey was

00:50:06.300 created by Scott Adams.

00:50:07.780 Not true.

00:50:09.280 So, Drew Carey, the TV show,

00:50:11.640 came on after Dilbert was

00:50:13.640 already kind of big.

00:50:14.960 And people said, hey, that guy

00:50:17.200 looks like Dilbert, and he works

00:50:20.780 in a cubicle, and must be

00:50:22.320 inspired by Dilbert.

00:50:24.240 Strange but true story.

00:50:26.800 Drew Carey called me home one

00:50:28.280 day, and he too, of course, had

00:50:32.020 been, you know, alerted to the

00:50:33.740 commonality.

00:50:35.200 So I got to actually chat to him

00:50:36.740 about his look.

00:50:40.080 And he had been wearing those

00:50:41.780 glasses for a long time, so that

00:50:43.180 was just his look.

00:50:43.940 So I had nothing to do with

00:50:45.880 Dilbert, it was actually a

00:50:47.280 coincidence.

00:50:48.560 And he offered me a job as a

00:50:51.000 writer on his show.

00:50:53.080 Now, it turns out that writers

00:50:55.620 are pretty well paid, but Dilbert

00:50:57.840 had already taken off by that

00:50:59.680 point.

00:51:00.460 So I didn't need a new job, so I

00:51:03.040 politely declined.

00:51:04.580 But no, if you think that one

00:51:06.520 copy of the other, it actually was

00:51:07.980 a coincidence.

00:51:09.080 So I can confirm that with

00:51:12.080 certainty.

00:51:13.520 Yeah.

00:51:14.920 He's a good guy.

00:51:16.140 I liked him.

00:51:19.920 And that, ladies and gentlemen,

00:51:22.400 concludes the 420 presentation.

00:51:24.900 Scott, just senior, 16,000 likes, okay?

00:51:36.420 Duke Carey is a libertarian, you

00:51:38.040 say.

00:51:38.440 Oh, okay.

00:51:42.100 Yeah, don't bring your martial arts

00:51:43.560 to a gunfight.

00:51:46.360 Yep.

00:51:46.760 So, uh, that's all I got for today.

00:51:50.960 Did I miss anything?

00:51:52.960 I think we can say it was the best

00:51:54.660 show you've ever seen so far.

00:51:55.920 Uh, DuckDuckGo runs on Bing.

00:52:02.980 Terrific.

00:52:07.020 Oh, yes, uh, I'll be on, that is

00:52:09.140 correct.

00:52:09.780 I will be on Dr.

00:52:11.080 Drew tonight.

00:52:12.660 6 p.m.

00:52:14.440 My time, so that's California time,

00:52:17.000 so adjust appropriately, or something

00:52:19.900 like that.

00:52:20.380 I think it's, I think that's what it is.

00:52:23.320 Um, yeah.

00:52:26.220 So, we'll see you there.

00:52:27.240 Maybe I'll see you there.

00:52:30.380 We'll tweet about it.

00:52:32.660 Ramaswamy versus Don Lemon.

00:52:34.260 Yeah, I saw a little clip of Don Lemon

00:52:36.900 and Ramaswamy trading words, but it

00:52:39.040 sounded like they were just talking

00:52:40.040 over each other, and I'm not sure

00:52:42.320 that that went anywhere.

00:52:44.220 Uh, no, there won't be a Trump RFK

00:52:46.340 Jr. ticket.

00:52:47.160 That's crazy.

00:52:48.960 That's crazy.

00:52:50.220 That's never gonna happen.

00:52:52.220 So, did you see Tucker Carlson

00:52:54.040 interview RFK Jr.?

00:52:56.540 That was a weird, sort of a weird

00:52:59.760 moment in politics, wasn't it?

00:53:02.140 Because RFK Jr. is a Democrat,

00:53:04.520 running as a Democrat, famous

00:53:06.380 Democrat family, but because of his

00:53:09.520 stance on vaccinations specifically,

00:53:12.800 I think, Tucker has a lot of

00:53:16.760 respect for him, and I would, I would

00:53:18.220 agree that he has earned our respect

00:53:21.880 and he's, should be taken quite

00:53:24.000 seriously in the race.

00:53:25.960 Now, did any of you take note of his

00:53:29.500 voice quality improvement?

00:53:33.320 Did you think he, his voice quality

00:53:35.400 was better?

00:53:38.480 Yeah, it looks like, it looked like

00:53:39.900 it improved, and I think it's still

00:53:41.720 improving, but I don't know.

00:53:45.000 Yeah, I believe it's still improving.

00:53:47.600 So, we'll see what the, uh, what the

00:53:50.560 upside on that is.

00:53:51.400 It might not be, I just don't know, but

00:53:53.360 I think it might be, because it feels

00:53:54.560 like it was better than the last time I

00:53:55.820 heard it.

00:53:58.280 Um, I feel like, I want to send him a

00:54:00.280 message to tell him there might be

00:54:02.820 something he could do with voice

00:54:04.180 production, uh, that he doesn't know

00:54:07.500 about.

00:54:07.800 Uh, specifically getting your voice up

00:54:10.580 into the mask of your, uh, face.

00:54:13.040 You might know how to do that, but

00:54:14.880 could be worth, uh, giving him a tip.

00:54:18.840 Because if you, if you bring your voice

00:54:21.120 production up here, I'll do it.

00:54:22.960 Uh, I will model for you what that looks

00:54:24.880 like.

00:54:25.120 If I bring my voice production up to the,

00:54:27.040 my face, I can actually feel my face

00:54:29.260 vibrating, as opposed to when I get lazy

00:54:32.260 and I talk down in my throat.

00:54:33.880 Now I'm talking to my throat.

00:54:35.900 Can you tell the difference?

00:54:37.300 It's almost guttural.

00:54:38.800 You can tell that my vocal cords and my

00:54:40.740 throat are the main part of my

00:54:41.940 production now.

00:54:43.160 But if I bring it up here, and the

00:54:45.900 humming is how you do it.

00:54:47.120 You hum.

00:54:49.020 And now, because I hummed, that kind of

00:54:52.040 allowed me to find how to bring the

00:54:54.020 production up at the top of my face.

00:54:56.040 And you notice that I don't even talk

00:54:57.760 nasally, and I'm not even boring.

00:55:01.740 Even though you say I am.

00:55:03.880 All right.

00:55:05.760 That's all for now, YouTube.

00:55:07.340 Thanks for joining.

00:55:09.360 Nice crowd.

00:55:10.600 Go forth and enjoy your 420.

00:55:13.260 And subscribe if you can.

00:55:15.280 And if you'd like to see the Dilbert

00:55:16.660 Reborn comic, and Robots Read News,

00:55:20.800 and lots of other stuff, go to

00:55:23.320 scottadams.locals.com for a

00:55:25.580 subscription.

00:55:26.800 Bye for now.

00:55:27.340 Bye for now.

Real Coffee with Scott Adams - April 20, 2023

Episode 2084 Scott Adams： Starship Launch, Hunter's Whistleblower, AI Bias, Can AI Be Hypnotized？

Episode Stats

Summary

Transcript