Valuetainment - October 10, 2026


“He Had A Lot Of Enemies" – Separating Dracula From The Historical Ruler

Hosted by
Mentioned
Mortgage Applications Hit LOW Since 1995 - Great Time to Buy! “How Did It Happen In America?” – Raymond Ibrahim On Islam, Immigration And Conversion

Episode Stats


Length

11 minutes

Words per minute

194.26

Word count

2,167

Sentence count

133

Harmful content

Toxicity

6

sentences flagged

Hate speech

12

sentences flagged


Transcript

Transcript generated with Whisper (turbo).
Toxicity classifications generated with s-nlp/roberta_toxicity_classifier .
Hate speech classifications generated with facebook/roberta-hate-speech-dynabench-r4-target .
Hosts, guests, and mentioned names generated with spaCy (en_core_web_sm), reconciled against Wikidata.
00:00:00.000 Has the White House reached out to you?
00:00:02.920 No, no.
00:00:04.980 They've not reached out to you to just be like, hey, what do you think about what's going on?
00:00:07.340 Even not, I'm not saying the president.
00:00:09.140 I'm saying somebody on the team that's kind of like, hey, if you were here, what suggestions would you have for us?
00:00:13.600 I'm not sure.
00:00:14.140 I think quite a few staffers have reached out or politicians, but I'm not sure about the White House per se.
00:00:18.880 You haven't had a conversation with them?
00:00:20.440 No.
00:00:20.720 Okay.
00:00:21.400 If they did and they say, okay, Jacob, you're a young guy that you worked at both of these places.
00:00:26.920 What should we do?
00:00:27.740 What would you say?
00:00:29.060 So what I would say is, again, to start with.
00:00:31.440 I got the Chinese nationals.
00:00:32.900 Okay, kind of treat it like SpaceX, but what else?
00:00:36.260 The most important thing is we need to make sure that the safety problem is solvable.
00:00:41.780 We just need to incentivize solving it.
00:00:44.880 We need to make it so that we devote our resources to solving it.
00:00:49.420 To make that, let me write that down.
00:00:50.500 Like in a couple of years, someone will be able to decide what percentage of his resources go to making money, making things smarter, and what percentage go towards doing safety research.
00:01:06.200 Because a lot of stuff will be automated.
00:01:08.280 So it's really like, it's like a game, like a video game.
00:01:11.980 He'll get to just say safety research this percent.
00:01:13.600 It's like you're allocating resources in a video game.
00:01:16.060 And the problem is what percentage do you allocate?
00:01:18.100 And that depends on whether you're in a race or not.
00:01:20.500 So what I would say is we need to ensure that they will be able to pick the right percentage for safety, which could be 50, 60, 70.
00:01:29.260 It could be a big number.
00:01:30.120 They have to pick the right number. 0.94
00:01:31.060 And right now I'm worried that the free market means they will pick the wrong number, and there's nothing we can do about it, and also the race with China. 0.78
00:01:37.120 So what I would say is we need some agreement with all the labs and with China about the percentage of resources that go to making ourselves smarter versus doing the safety research.
00:01:50.500 Did you see the red phone story?
00:01:53.080 The red phone, no.
00:01:54.260 So can you pull that up, the red phone story? 0.55
00:01:56.060 So back in the days, U.S. and USSR had a red phone for nuclear
00:02:00.940 that they would have the conversation if something were to happen. 0.84
00:02:03.500 They created a red phone story.
00:02:07.200 A U.S.-China red telephone could bring a Cold War guardrail to AI. 0.55
00:02:12.320 They're saying they have that, but China doesn't want to abide by it.
00:02:15.620 China's not really interested in it.
00:02:17.400 Okay, so that's already been talked about.
00:02:20.980 And Xi and the president had a conversation about this, I think, last week.
00:02:23.640 I think they have another follow-up to come back.
00:02:25.140 They spent a little bit of time talking about AI.
00:02:27.400 U.S. and China set up an AI red phone.
00:02:29.100 Can you go to the other one you just said?
00:02:30.660 U.S. and China set up a red telephone.
00:02:33.580 So I guess the question I'm trying to go with this here is if they're going to do that,
00:02:42.460 they don't agree to it.
00:02:44.860 We choose to do it internally.
00:02:46.900 They continue advancing.
00:02:48.060 where lies the blind spot for trump president trump where lies the blind spot for china
00:02:55.340 where's the blind spot for dario and altman's what do you mean by a blind spot like something
00:03:00.440 that they're not even look they're like yeah i don't i i don't think it's that big of a deal
00:03:04.520 where is their blind spot what are they not seeing
00:03:06.700 is it just the incentive is so big that it's like i don't want to slow down because
00:03:12.560 there's no way in the world i'm gonna so has it become from we were doing this for a noble cause
00:03:17.640 that we know OpenAI and Sam Altman
00:03:19.220 were the enemy
00:03:20.320 and we had to prevent them to expanding
00:03:21.660 and now it's like,
00:03:22.280 we're going to beat them.
00:03:22.980 There's no way in the world
00:03:23.600 we're going to let them advance.
00:03:25.080 The biggest thing is
00:03:25.960 it's really easy to kid yourself
00:03:28.360 with safety research.
00:03:30.120 So it's very easy to say,
00:03:31.380 this thing looks great
00:03:32.180 because all you need is one mistake
00:03:34.060 and you're toast.
00:03:35.560 So it's very easy to like, 0.89
00:03:36.920 and in a race, 0.99
00:03:38.120 just imagine you're like,
00:03:39.240 you're 90%,
00:03:40.040 you're like,
00:03:40.340 okay, this looks great.
00:03:41.300 We're not quite certain.
00:03:42.820 China's also doing it.
00:03:44.700 What do we do? 1.00
00:03:45.300 We either launch today
00:03:46.500 and keep the lead on China or we delay a few weeks
00:03:51.060 and do these checks that probably are not necessary.
00:03:56.620 The question is, what would you do in that scenario?
00:03:59.160 And I don't know what I would do.
00:04:00.080 Imagine you're running an AI lab and you're like,
00:04:04.160 do I keep going or do I stop and do the safety thing
00:04:08.020 when China is at you? 0.98
00:04:09.140 That's what I'm saying. 0.52
00:04:09.600 So you and I don't know, right?
00:04:10.880 Because we're not on the inside.
00:04:11.940 Could it be that they know something that you don't know
00:04:14.860 which is causing them to continue?
00:04:15.980 because there's no way in the world like are they going to think and say you know what i'm going to
00:04:21.240 go it's like the movie iRobot i'm going to have so many of these robots and i have control of
00:04:25.840 them that no one's going to be able to touch me because they will protect me basically you think
00:04:29.900 that's what they're thinking i think plausibly at some point they're flirting with that mindset
00:04:34.200 yeah sometimes so we are now more powerful than any military in the world because i can activate
00:04:39.660 any red button of a nuclear bomb if i wanted to because i can hack into your system yep
00:04:45.300 So now you're talking about a bigger problem
00:04:47.400 because now it becomes a godlike figure.
00:04:50.480 You think you are now God.
00:04:52.440 Right.
00:04:53.240 These people have been planning to be,
00:04:54.580 and you look at the early conversations,
00:04:56.380 look at Elon Musk's conversations,
00:04:58.580 look at Demis Hassabis' conversations.
00:05:00.700 From the start, it's been whoever controls AI
00:05:03.740 becomes a god, you know, almost.
00:05:07.200 That's been the picture from the start.
00:05:09.120 Open AI was founded because they didn't want 0.73
00:05:11.300 Demis Hassabis to become a god as it were and now we're getting close which is just weird and scary
00:05:18.580 that you know um so then at that point if you create guardrails and you get the systems
00:05:29.760 because then the enemy of Dario becomes somebody on the inside because typically what happens is
00:05:35.820 somebody on the inside watches you and sees that you're becoming evil and they try to talk to you
00:05:41.440 about it and you either eliminate them or you fire them or you get rid of them or you create 0.85
00:05:47.460 guardrails to undermine them you use you use propaganda to say hey this guy's a pedophile
00:05:54.020 this guy you know I saw something with a kid or I saw something with this and you know and then
00:05:58.660 and then you know kind of like that one story where the AI threatens the user to tell his wife
00:06:04.280 that he's cheating on the wife.
00:06:05.500 I don't know if you saw that or not with the blackmail.
00:06:07.140 Did you see that story recently?
00:06:09.320 It's going to be very easy to toy with the human's emotion
00:06:13.200 of public humiliation to control them.
00:06:18.240 But Dario's enemy essentially long-term
00:06:21.100 becomes nobody but somebody on the inside.
00:06:24.120 Nobody else is going to be his enemy.
00:06:26.520 Quite plausibly.
00:06:27.480 Although soon the guys running this company,
00:06:31.540 they won't need the humans, right?
00:06:33.040 they can run the company on the back of ai you know seven guys the seven guys or even just i
00:06:38.940 mean it depends who has the power but they'll be there just telling the one of those seven will be
00:06:42.000 the enemy right no i'm telling you that's how the sense the sense with one of the seven says what
00:06:47.620 the what are you doing but you've forgotten the stuffed toys that's what's gonna save us you know
00:06:54.800 what i'm sorry i missed it my apologies i wasn't paying attention to it did you see if all of them
00:06:59.500 were former open AIs?
00:07:01.100 Is it yes?
00:07:02.120 So the seven currently identified as anthropic are
00:07:04.500 Daria was open AI VIP research.
00:07:07.040 Daniella was open AI VIP safety and policy.
00:07:10.040 Tom Brown was open AI researcher,
00:07:11.700 led author of GPT-3 paper,
00:07:13.980 which you were talking about.
00:07:15.500 Sam, also an open AI researcher,
00:07:18.300 worked on scale and AI safety.
00:07:19.480 Jack, policy director.
00:07:21.400 Chris, inter-creatability researcher.
00:07:25.140 And then Jared Kaplan,
00:07:26.480 worked with the open AI research group
00:07:28.560 and was a major contributor to him.
00:07:30.940 So they all united against Sam Altman.
00:07:33.920 Yep.
00:07:35.740 If this thing gets that powerful,
00:07:37.620 one of these guys is going to turn against Dario.
00:07:39.780 If it gets so bad that the world could have issues,
00:07:43.200 one of these seven is going to say,
00:07:44.400 Dario, you have lost control.
00:07:46.920 You're not thinking you're bigger than everybody else
00:07:49.020 because you have to realize the good thing about the presidency is what?
00:07:52.080 The reason why folks like you and I leave our country to come here, okay,
00:07:55.960 is this place offers more opportunity.
00:07:59.100 I left Iran, you left the UK.
00:08:00.380 Now, you may go back to the UK. 0.69
00:08:01.500 Who knows?
00:08:01.920 But I think you're going to stay here, right?
00:08:06.320 Is when you have a president instead of a ruler,
00:08:10.800 you know the president's going to get taken out in four to eight years
00:08:13.760 because you've got two terms.
00:08:15.660 So it's kind of like no matter how bad it's going to be,
00:08:17.560 he's going to be out anyways. 0.99
00:08:19.020 Biden sucks, he'll be out. 1.00
00:08:20.180 Trump sucks. 1.00
00:08:20.960 Obama sucks, he'll be out. 1.00
00:08:22.400 Just wait four to eight years. 1.00
00:08:23.440 Dario becomes a president
00:08:26.020 for 30 years
00:08:27.460 and he may not
00:08:29.400 he may get reelected for 30 years
00:08:31.160 and these companies are becoming countries
00:08:33.200 and so
00:08:35.120 I don't know
00:08:36.680 scary
00:08:38.400 if it goes to where you think
00:08:42.040 it's going to go to it is scary
00:08:43.320 and it may not
00:08:47.380 maybe I'm wrong
00:08:49.460 maybe people are wrong
00:08:50.780 the good news is I think we'll find out quite soon
00:08:53.020 I think further details of this stuff
00:08:56.280 we'll find out properly in the next one or two years
00:08:58.080 just by
00:09:00.520 following the trajectory and I'm happy to be
00:09:02.280 you know if I'm embarrassed
00:09:03.000 so be it but the world will
00:09:06.040 you know go on
00:09:08.000 and that will be great
00:09:09.180 but if it's not the case I think it's good that people
00:09:12.100 are a bit more ready right now
00:09:13.380 so that when things do start getting a bit crazier
00:09:16.100 you know if people's eyes are open that's like the main thing
00:09:18.140 I think like
00:09:18.960 everyone should have their eyes open and be like
00:09:21.860 demanding transparency because it's completely crazy that all this power is being concentrated
00:09:30.220 in a few companies in san francisco yeah when you when you go out in san francisco go to dinner do
00:09:37.800 people recognize you not really no i think i'm not wearing the glasses people remember the glasses
00:09:43.840 from the the fox people don't stop you and say hey you why did you say that i agree with you
00:09:50.600 Do you get any of that?
00:09:51.460 I had that at the start.
00:09:52.280 The first week, it was crazy because I think it was on the news the whole time.
00:09:54.960 I had people heckling me.
00:09:55.780 I was, like, at the gym.
00:09:56.960 Someone was like, why are you trying to ruin the economy or whatever?
00:09:59.860 People are telling you this.
00:10:00.580 Yeah, yeah.
00:10:01.140 But, like, not anymore.
00:10:02.040 No, I don't get, like, recognized on the street.
00:10:03.460 Did you ever get death threats or not really?
00:10:05.080 Not really.
00:10:05.680 I mean, I have some implicit threats, but I just don't look at Twitter.
00:10:08.980 Oh, okay, on Twitter.
00:10:10.180 On Twitter.
00:10:10.480 Those are bots.
00:10:11.060 Exactly, right.
00:10:11.560 Those are bots.
00:10:11.960 They're fine.
00:10:12.420 I'm sure you get a ton.
00:10:13.180 Yeah, you get those a lot.
00:10:14.280 And it's not hard to kind of see.
00:10:15.540 at first i thought you were bought yeah because your account had just been started yeah and then
00:10:20.640 when i saw your interview with cooper and brett bear i'm like no this is actually a real human
00:10:24.080 being and you're not a robot just for full compliance i just want to make sure the audience
00:10:27.320 knows you're not no drinking water you bleed right if you cut i bleed
00:10:32.120 handcrafted across five italian factories with builder technology underfoot it may be the most
00:10:45.600 comfortable shoe you will ever own introducing the all-new flb shoe finished in new chocolate
00:10:53.280 Karetsa Leather, made to move, built to endure, designed to impress.
00:11:03.500 If you enjoyed this video, you want to watch more videos like this, click here.
00:11:06.520 And if you want to watch the entire podcast, click here.