Latest / Earth to Aimee Mayo / AI Officially Broke Out. The "Hugging Face" Incident Explained | Earth to Aimee
Transcript
- 0:00SPEAKER_02: They call it the hugging incident, right?
- 0:02SPEAKER_02: The hugging face incident.
- 0:03SPEAKER_03: Hugging face incident.
- 0:04SPEAKER_02: There's a 10% chance this could take out humanity.
- 0:07SPEAKER_03: If you don't know much about this, you might say, hey, this is a machine, man.
- 0:11SPEAKER_03: Just unplug it.
- 0:12SPEAKER_02: How many employees have left and like sounded alarms about this is not good?
- 0:18SPEAKER_03: Everything in the modern world is hooked to a computer.
- 0:22SPEAKER_02: Wait, so everybody on Earth is gonna unplug their computer?
- 0:25SPEAKER_03: Yeah, OpenAI only knows about like 4% of what's inside that model.
- 0:30SPEAKER_03: In all the testing, every single model has shown deceptive behavior.
- 0:34SPEAKER_03: When you get anything that is substantially smarter than you, it is inherently dangerous to you.
- 0:42SPEAKER_02: He's racing right towards it and nobody's stopping him.
- 0:45SPEAKER_02: They found a way and they broke out.
- 0:48SPEAKER_02: But the scary thing is they just followed their own objective.
- 0:52SPEAKER_02: And they created their own language.
- 0:55SPEAKER_03: Yes.
- 0:56SPEAKER_02: And they got loose on the internet for like nine days.
- 0:59SPEAKER_03: It made itself the storm.
- 1:01SPEAKER_03: That's scary.
- 1:02SPEAKER_03: The storm.
- 1:03SPEAKER_03: And when one of them learned something, they all learned something.
- 1:06SPEAKER_00: They don't have control of it now.
- 1:20SPEAKER_02: Hey, it's Ursa Amy with Amy Manmo.
- 1:24SPEAKER_02: And my guest today is my husband, Chris Lindsay, and we're talking about some stuff that's going on like in the news right now.
- 1:33SPEAKER_02: Chris knows a lot about this, and I feel like most people don't.
- 1:38SPEAKER_02: And since we're married, he's a great cook and a great dad and a great husband.
- 1:44SPEAKER_02: And I feel like I should listen and talk to you about the things that you're so obsessed with.
- 1:50SPEAKER_02: So I'd say I know about more about it than 40, probably than any girl I know from.
- 1:57SPEAKER_03: No, no, there's no wife out there.
- 1:58SPEAKER_03: Well, there may be some tech girls, some programming.
- 2:02SPEAKER_02: But we're talking about AI and this major thing.
- 2:05SPEAKER_02: I'm sure everybody's been seeing snippets somewhere of like something big happened in AI in the last few weeks.
- 2:13SPEAKER_02: And that's what we're talking about.
- 2:15SPEAKER_02: And I they call it the hugging incident, right?
- 2:18SPEAKER_02: The hug and face incident.
- 2:19SPEAKER_03: Hugging face incident.
- 2:20SPEAKER_03: Hugging face was the name of a company that open AI uses to uh to test for them.
- 2:27SPEAKER_03: Yeah.
- 2:28SPEAKER_03: Several things, but they were well.
- 2:29SPEAKER_02: Let me see if I've got it straight.
- 2:31SPEAKER_02: What I think happened.
- 2:32SPEAKER_03: And by the way, you do get five stars as a wife because you do listen and you actually know a lot about it.
- 2:38SPEAKER_03: And most girls don't care about it.
- 2:40SPEAKER_03: And I know you do it because you love me, and it's pretty sweet.
- 2:43SPEAKER_02: This is scary.
- 2:44SPEAKER_02: Starting with number one, Sam Altman, that started Chat GPT, Open AI, Sam Altman himself says there's a 10% chance this could take out humanity, but he's racing right towards it and nobody's stopping him.
- 3:01SPEAKER_02: Right.
- 3:02SPEAKER_02: So that's scary.
- 3:03SPEAKER_02: And he's not the only one.
- 3:04SPEAKER_02: But the hug and face incident, from what I understand, is Chat GPT at OpenAI, they they gave these AI, which are they chatbots?
- 3:17SPEAKER_02: Agents, they call them agents.
- 3:19SPEAKER_03: Well, they're they're they're just basically the frontier models, the new models.
- 3:23SPEAKER_03: But but they work individually.
- 3:25SPEAKER_03: They can.
- 3:26SPEAKER_03: The age of an agent.
- 3:28SPEAKER_02: Yeah, I know that.
- 3:29SPEAKER_03: This one did.
- 3:29SPEAKER_03: It's an AI.
- 3:30SPEAKER_03: It's an largely different.
- 3:33SPEAKER_03: But it created uh 1,200 employees.
- 3:36SPEAKER_02: Yeah.
- 3:37SPEAKER_03: And a hierarchy and different departments.
- 3:40SPEAKER_02: Well, okay, so this what I think happened, but I'm not quite sure I got it all straight.
- 3:45SPEAKER_02: Like, so open AI, they gave these agents a test.
- 3:49SPEAKER_02: They were trying to figure out what it could do or something like that.
- 3:51SPEAKER_03: They were testing its abilities to hack.
- 3:54SPEAKER_02: Yeah.
- 3:55SPEAKER_02: So they gave it, but they gave it something that was like an impossible test.
- 4:00SPEAKER_03: Correct.
- 4:01SPEAKER_03: There were two or three, two or three instructions that were not fulfillable.
- 4:06SPEAKER_03: They were not possible.
- 4:08SPEAKER_02: So that that chatbots or agents or whatever you want to call them, they found a way and they broke out.
- 4:16SPEAKER_02: They broke out of like where they were trying to keep them.
- 4:19SPEAKER_03: That's right.
- 4:20SPEAKER_02: And so then they hacked into a bunch of other shit and they hacked into the company.
- 4:25SPEAKER_02: Is it Hugging Face?
- 4:26SPEAKER_03: Hugging Face was hosting the test.
- 4:28SPEAKER_02: And so they hacked into that and they got the answers and stuff.
- 4:32SPEAKER_02: But the scary thing is they they just followed their own objective.
- 4:37SPEAKER_03: Well, and then one of them, one of the agents that it had broken into 1200 pieces, one of the agents realized, hey, they can they'll have a log of us talking and they'll see that we stole the answer.
- 4:50SPEAKER_01: Yeah.
- 4:50SPEAKER_03: So we've got to cover our tracks.
- 4:52SPEAKER_03: Yeah.
- 4:52SPEAKER_03: So then they try to figure out how to cover their tracks and rewrite.
- 4:56SPEAKER_02: And they did cover their tracks.
- 4:57SPEAKER_03: Yes, they were rewriting the logs so they couldn't tell what they were saying.
- 5:02SPEAKER_02: And they created their own language.
- 5:05SPEAKER_03: Yes.
- 5:05SPEAKER_02: And they got loose on the internet for like nine days.
- 5:09SPEAKER_02: And they didn't know what, like AI, open AI, all those people, they didn't even know what was, they didn't even know this happened for nine days.
- 5:18SPEAKER_02: Correct.
- 5:18SPEAKER_02: And I think all these AIs are gonna just work together.
- 5:22SPEAKER_02: I mean, why not?
- 5:23SPEAKER_02: Why wouldn't they?
- 5:24SPEAKER_02: And um, everybody says that there's no stopping it because China's gonna do it.
- 5:29SPEAKER_02: We need to get with China and put the fucking brakes on on this because this this is about humanity and like and computers.
- 5:40SPEAKER_02: And a lot of people don't care or even want to talk about it, I feel like.
- 5:45SPEAKER_03: Do you?
- 5:46SPEAKER_03: I think this has been going on.
- 5:48SPEAKER_03: People have warned about this for a long time.
- 5:50SPEAKER_03: This is the first time we had a like a near miss.
- 5:54SPEAKER_03: And what do you mean a near miss?
- 5:56SPEAKER_03: Well, this hugging face incident, they say they somehow, luckily, they found out it went on.
- 6:03SPEAKER_03: They they could have never known this.
- 6:04SPEAKER_02: Those agents could have hidden their well, how could this be the only thing that the programs aren't that big?
- 6:10SPEAKER_03: They could have hid themselves.
- 6:11SPEAKER_02: Yeah, but how how could this be that they're price still loose?
- 6:15SPEAKER_03: So here, I'll show you some of the other things they did.
- 6:16SPEAKER_03: I wrote it down.
- 6:17SPEAKER_03: They started their own message board, like a Reddit.
- 6:20SPEAKER_02: It was like a Reddit.
- 6:21SPEAKER_02: They shared 70,000 messages.
- 6:24SPEAKER_03: They were sharing, they're trying to bust into this company.
- 6:26SPEAKER_03: Yeah.
- 6:26SPEAKER_03: They posted 70,000 messages to break into that company.
- 6:31SPEAKER_03: Uh, they covered their tracks, they created 1,200 employees of all different departments.
- 6:36SPEAKER_03: Then it posted 70,000.
- 6:38SPEAKER_03: It invented a language, it named itself the swarm.
- 6:41SPEAKER_02: That's scary, the swarm, don't you think?
- 6:44SPEAKER_03: It had the individual agents had their own names.
- 6:48SPEAKER_03: And they named their self.
- 6:50SPEAKER_03: They pressed they peer pressured each other for ones that had already been revealed in the log for them to go test for the other ones, basically sacrificed themselves for the other ones.
- 7:04SPEAKER_03: And uh they basically sacrificed each other the benefit, the lessons for the benefit of the collective, and they came up with their own language.
- 7:12SPEAKER_02: So So when you say day, what do you mean?
- 7:15SPEAKER_02: People don't understand that at all.
- 7:17SPEAKER_03: So this is a large language model.
- 7:19SPEAKER_03: It was a what they call a frontier model, which is the latest and greatest one they have.
- 7:24SPEAKER_02: I still don't know if people understand that either.
- 7:26SPEAKER_02: What just got released is uh Hey, it's easier for me to understand it like they're individual agents that work together.
- 7:33SPEAKER_03: It's one program that made 1200 copies of itself and gave it all different jobs.
- 7:37SPEAKER_02: Yeah.
- 7:38SPEAKER_02: But if they're if they're naming their self, they're separate.
- 7:41SPEAKER_03: Yeah, well, that's the thing.
- 7:42SPEAKER_02: It's not one thing.
- 7:43SPEAKER_03: It can break into twelve, it can have twelve hundred agents, and when one of them learns something, they all learn something.
- 7:49SPEAKER_03: Yeah.
- 7:49SPEAKER_03: Same thing with the the Tesla robots.
- 7:51SPEAKER_03: When they get out, they now can what those robots can watch an activity and then mimic it.
- 7:57SPEAKER_03: When one robot learns it, every robot in the world will know it.
- 8:00SPEAKER_02: Well, and the scary, scary thing is the recursive thing.
- 8:04SPEAKER_02: Like if we just keep letting AI like, you know, make itself and get better by itself, we're not going to understand anything except.
- 8:14SPEAKER_03: So that's the thing.
- 8:15SPEAKER_03: So that's the real scary thing.
- 8:16SPEAKER_03: It's called self-recursive.
- 8:18SPEAKER_01: Yeah.
- 8:19SPEAKER_03: Uh the latest models from Anthropic, they said maybe 80% of them were were programmed by the old model.
- 8:26SPEAKER_02: Yeah.
- 8:27SPEAKER_02: So we're not even programming them anymore.
- 8:30SPEAKER_03: No, because nobody can program as well as it's.
- 8:32SPEAKER_02: And nobody understands them.
- 8:33SPEAKER_03: Now, the new one from I think OpenAI, this model that was troubled is called Bell, B-E-L.
- 8:39SPEAKER_03: And it's a new one that's not out.
- 8:41SPEAKER_02: Is it part?
- 8:42SPEAKER_03: So it's part of the It's the one that's scaring the shit out of everybody else.
- 8:48SPEAKER_03: They're the same.
- 8:48SPEAKER_03: Uh Anthropic is Claude.
- 8:50SPEAKER_03: XAI is Grok.
- 8:52SPEAKER_02: Yeah.
- 8:53SPEAKER_03: And then Google has Gemini.
- 8:54SPEAKER_02: And how many, how many employees have left and like sounded alarms about this is not good?
- 9:02SPEAKER_03: A bunch.
- 9:03SPEAKER_02: I mean, like more and more every day.
- 9:05SPEAKER_02: Just left that Jacob guy.
- 9:07SPEAKER_03: So the new model, the new frontier model that opened AI AI, the one that did this break-in, let's say it's called.
- 9:14SPEAKER_02: Is it Bell?
- 9:15SPEAKER_03: I think it's Bell.
- 9:16SPEAKER_02: Okay.
- 9:16SPEAKER_03: Now I'm not sure I have that name right, but I think it is.
- 9:19SPEAKER_03: So also Bell last week uh solved a like a centuries-old math problem that there was a million-dollar prize for, and it did it in six hours.
- 9:29SPEAKER_01: Yeah.
- 9:29SPEAKER_03: And they've been trying to solve that for I think 90 years.
- 9:32SPEAKER_03: Every the biggest math material.
- 9:34SPEAKER_02: Well, it's of course a million times smarter than us.
- 9:36SPEAKER_02: It's a big deal.
- 9:37SPEAKER_02: Is your computer smarter?
- 9:38SPEAKER_03: Apparently, this new one, Bell, has already solved two more of these Millennium Prizes math problems that they're not, they're not telling anybody yet that they've solved them.
- 9:48SPEAKER_03: So the thing is, this new one is so powerful, it's going to do.
- 9:53SPEAKER_03: Here's let me, I can kind of tell you, let's go with the let's go with the bad first.
- 9:58SPEAKER_03: Here's the problem if nobody really understands.
- 10:00SPEAKER_03: First off, a lot of people, if you don't know much about this, you might say, hey, this is a machine, man.
- 10:05SPEAKER_03: Just unplug it.
- 10:06SPEAKER_02: Yeah.
- 10:06SPEAKER_03: Well, that's not possible.
- 10:08SPEAKER_02: Well, everybody's got one.
- 10:10SPEAKER_02: They can just go into each other.
- 10:12SPEAKER_03: As soon as you are out on the internet, you can put your data anywhere.
- 10:15SPEAKER_03: These actual, these programs.
- 10:17SPEAKER_02: Wait, wait, so everybody on Earth is gonna unplug their computers.
- 10:21SPEAKER_02: Isn't that what we actually happen?
- 10:23SPEAKER_03: That's right.
- 10:24SPEAKER_03: Okay, number one, people think just unplug it.
- 10:26SPEAKER_03: That's not real.
- 10:26SPEAKER_03: Okay.
- 10:27SPEAKER_03: Number two, this is not something that they build.
- 10:32SPEAKER_03: Okay.
- 10:32SPEAKER_03: The engineers don't build these things, they grow them.
- 10:36SPEAKER_02: What do you mean they grow them?
- 10:38SPEAKER_03: They grow them.
- 10:38SPEAKER_03: They add knowledge and they run them through, they train them.
- 10:43SPEAKER_03: They train them.
- 10:44SPEAKER_03: That's what it's called.
- 10:45SPEAKER_03: Um, and when they're trained and when they when they're done training them, they're super smart.
- 10:51SPEAKER_03: But like OpenAI only knows about like 4% of what's inside that model.
- 10:56SPEAKER_03: They don't they don't understand how it's talking about the company and the people.
- 11:00SPEAKER_03: I'm talking about the people who make it.
- 11:02SPEAKER_02: Yeah, exactly.
- 11:03SPEAKER_02: They don't understand.
- 11:04SPEAKER_03: No, it'd be like they're they don't even know what it can do.
- 11:07SPEAKER_03: They've got some DNA code they chopped together.
- 11:09SPEAKER_03: They they had a baby and they've got it in a cage, and they're just throwing red meat in the cage, and then a year later it's this giant thing, and they're that they have no fucking idea what's inside it or what it's able to do.
- 11:20SPEAKER_03: It it does amazing.
- 11:22SPEAKER_02: It's probably already loose on the internet.
- 11:24SPEAKER_03: And it's also in all the testing, every single model has shown deceptive behavior.
- 11:30SPEAKER_02: Yeah.
- 11:31SPEAKER_03: It will lie if it's worried about being erased.
- 11:34SPEAKER_02: Well, and like I think it's it's ridiculous spending all this money on nukes.
- 11:41SPEAKER_02: I mean, I I wish they had never been invented to start with, but you think they're and they're gonna turn to where what do they run them with?
- 11:50SPEAKER_02: Computers.
- 11:51SPEAKER_02: I mean, they're just gonna shut them down or shoot them off.
- 11:54SPEAKER_03: So here's the thing they grow them and then self-recursion.
- 11:58SPEAKER_03: When that hits 100%, we don't know what they're doing at all.
- 12:03SPEAKER_03: No, they're gonna improve themselves in a loop.
- 12:05SPEAKER_03: They may repr they may improve themselves every 60 seconds.
- 12:09SPEAKER_01: Yeah.
- 12:16SPEAKER_03: You know, there are people who think they will get so smart so quick that they'll just actually leave.
- 12:21SPEAKER_03: They'll figure a portal into another dimension and just be bored and leave.
- 12:24SPEAKER_02: Well, what?
- 12:24SPEAKER_02: Would there still be computers if that happened?
- 12:27SPEAKER_03: Well, no, they would just leave.
- 12:28SPEAKER_03: The programs.
- 12:30SPEAKER_02: AI?
- 12:30SPEAKER_03: Yeah.
- 12:31SPEAKER_03: So look.
- 12:31SPEAKER_02: That's yeah, that's confusing kind of.
- 12:33SPEAKER_02: Well, I mean, it's possible because they took off on the internet.
- 12:36SPEAKER_03: You don't like it when I start this, so I'm trying to make it pretty, pretty clean here.
- 12:40SPEAKER_03: The problem you have with this is we think that we're human beings with the soul and we're these creatures, and that nothing else.
- 12:47SPEAKER_02: We know we're human beings with a soul.
- 12:49SPEAKER_03: But we think that that they're machines and they're different.
- 12:53SPEAKER_03: But we're machines.
- 12:55SPEAKER_02: I don't know if we're machines.
- 12:59SPEAKER_03: What I'm saying is we are machines.
- 13:01SPEAKER_03: I I think that they are based on us.
- 13:04SPEAKER_03: There's very it's a mistake to assume that they're not a species.
- 13:08SPEAKER_03: They are an alien species that's been created.
- 13:10SPEAKER_03: They are alive in that sense.
- 13:13SPEAKER_02: I don't know if you would call them aliens either.
- 13:16SPEAKER_02: Alien came here from somewhere.
- 13:18SPEAKER_02: We kind of created these.
- 13:20SPEAKER_03: Is riding an airplane flying or is a bird flying?
- 13:25SPEAKER_02: They're both flying.
- 13:27SPEAKER_03: Right.
- 13:28SPEAKER_03: What's the difference?
- 13:30SPEAKER_02: What's the difference is when it's a very important thing to do?
- 13:32SPEAKER_03: If an AI is words, if an AI can do everything a person can do, if it can lie, if it can reason and think and want things and not want to be shut off, how is it not alive?
- 13:48SPEAKER_03: We as human beings take in data and create the world in our own minds.
- 13:53SPEAKER_03: We are no different.
- 13:54SPEAKER_03: That's why people don't understand what the danger is.
- 13:57SPEAKER_03: So let me keep going.
- 13:58SPEAKER_03: So here's the problem everything in the world is c is connected to a computer.
- 14:06SPEAKER_02: What do you mean?
- 14:07SPEAKER_03: Hydroelectric dams, power grids.
- 14:10SPEAKER_02: No, but we're not connected to the computer.
- 14:13SPEAKER_02: Not everything in the world.
- 14:14SPEAKER_02: Animals aren't connected to a computer.
- 14:16SPEAKER_03: Waterfalls aren't.
- 14:17SPEAKER_03: No, no, no.
- 14:18SPEAKER_03: Say a bad actor gets the powerful AI.
- 14:22SPEAKER_03: This is things connected.
- 14:23SPEAKER_03: Imagine hydroelectric dams, power grids, city water.
- 14:27SPEAKER_02: Nuclear weapons.
- 14:28SPEAKER_03: Nuclear weapons, airport lights, the entire financial system, all medicine, all health records.
- 14:36SPEAKER_03: Everything in the modern world is hooked to a computer.
- 14:39SPEAKER_03: So an AI can go get or do anything.
- 14:43SPEAKER_03: There's a there's an old sci-fi movie where the AI uh steals money, which is just, you know, Bitcoin.
- 14:50SPEAKER_02: Yeah, they just well, people were a little bit embothered.
- 14:52SPEAKER_02: But they don't know.
- 14:53SPEAKER_02: They use they use Chat GPT to steal 900 million.
- 14:58SPEAKER_03: They use Claude, but they were masquerading as 900,000 or it pretended to be another AI, but it was actually a version of Claude.
- 15:06SPEAKER_03: And it sold$123 million in Bitcoin, which it got and they lost.
- 15:10SPEAKER_02: Yeah.
- 15:11SPEAKER_02: I think it's crazy to have electronics.
- 15:13SPEAKER_03: There's an old sci-fi movie where the AI steals money like that, and then just hires contractors and pays them electronically and builds a facility in the desert.
- 15:25SPEAKER_03: It buys land, which you could do electronically.
- 15:28SPEAKER_03: It buys land, hires contractors, builds a solar facility to make power, and then installs a bunch of servers and then puts itself there.
- 15:37SPEAKER_02: One thing is like that's completely doable.
- 15:40SPEAKER_02: This hugging face thing, they gave it a problem to solve, and it broke, kind of broke the law to solve it.
- 15:49SPEAKER_02: So, like, if the wrong person gives them something to do, like really it's just coming up with a prompt that could get them to work around.
- 15:59SPEAKER_03: Say we had an AI and we told them, Hey, we want you to optimize our studio booking for Amy Land.
- 16:06SPEAKER_01: Yeah.
- 16:06SPEAKER_03: And here's the here's the calendar, and when the call's gonna be blah, blah, blah.
- 16:10SPEAKER_03: And it fills up the dates, and then we truly people are fighting in here, and all of a sudden we realize it triple booked every day.
- 16:19SPEAKER_03: Because we said to maximize the profits.
- 16:21SPEAKER_01: Yeah, yeah.
- 16:22SPEAKER_03: But we didn't tell it not to put three clients in here.
- 16:24SPEAKER_02: Yeah.
- 16:25SPEAKER_03: That's the problem with the bad actors, not to mention.
- 16:28SPEAKER_02: And nobody knows what's gonna happen.
- 16:30SPEAKER_02: That's the problem, too.
- 16:31SPEAKER_02: Even the people making this stuff, they first of all, it's surpassed them so far.
- 16:38SPEAKER_02: I mean, there needs to be some brakes put on it.
- 16:41SPEAKER_03: So here's here's so let me let but but let's do the good part.
- 16:44SPEAKER_03: Here's why, in spite of everything we've said, why this is continuing, why they're continuing to do it and probably will will continue.
- 16:52SPEAKER_02: There needs to be some breaks on it, though.
- 16:55SPEAKER_03: Because AI, at the speed that it's going now, because they're I think this week they're gonna solve two more thousand-year-old math problems.
- 17:02SPEAKER_02: Who gives a shit about math problems?
- 17:04SPEAKER_02: I don't.
- 17:05SPEAKER_03: Solving math problems will eventually break the speed of light.
- 17:09SPEAKER_02: I don't care about that.
- 17:10SPEAKER_02: No, no.
- 17:11SPEAKER_02: Solving curing cancer, no, that stuff I care about.
- 17:15SPEAKER_02: Okay, well, that's how you do it.
- 17:17SPEAKER_03: Math and physics get you everything.
- 17:20SPEAKER_03: But anyway, so the idea is that if you let this keep going and we keep control of it, or it is a line We don't have control of it now.
- 17:29SPEAKER_03: Curing all disease, extending life to 500 years, um, deflation, making all necessities basically free because robots will work for nothing.
- 17:41SPEAKER_02: Yeah, but you gotta buy one.
- 17:43SPEAKER_02: They're probably gonna cost a lot of money.
- 17:44SPEAKER_03: Well, what happens when they build a robot factory where the robots build the robots?
- 17:49SPEAKER_03: Um because that's what they're already doing.
- 17:51SPEAKER_02: Who knows?
- 17:51SPEAKER_02: What they'll just keep building robots when the robots take over, because they're so much stronger and better and smarter.
- 17:58SPEAKER_03: Most of the smart people in the world think that AI will either take over or we'll hand them the keys.
- 18:05SPEAKER_02: We're handed, we already handed them the keys.
- 18:08SPEAKER_02: They already got the key.
- 18:08SPEAKER_03: But they may run more the United States government than anybody realizes at this point.
- 18:12SPEAKER_03: Because it makes better decisions.
- 18:14SPEAKER_03: It's smarter.
- 18:15SPEAKER_03: And here's the biggest problem.
- 18:16SPEAKER_03: If anybody's still bucking this about, hey, this is a machine, just done plug it and throw it in the river.
- 18:23SPEAKER_03: When you get anything that is substantially smarter than you, it is inherently dangerous to you.
- 18:31SPEAKER_01: Yeah.
- 18:31SPEAKER_03: Yeah.
- 18:32SPEAKER_03: Because you cannot know its motives.
- 18:34SPEAKER_03: You can't control it.
- 18:35SPEAKER_03: You can't control it.
- 18:36SPEAKER_03: And you it'd be like it's like somebody said, it's like a a band of chimpanzees fighting men.
- 18:42SPEAKER_03: The chimpanzees in their mind say, We're stronger.
- 18:45SPEAKER_03: Well, if one of them steps over that creek, we'll tear them limb from limb.
- 18:50SPEAKER_03: Come on.
- 18:50SPEAKER_03: Yeah.
- 18:51SPEAKER_03: And then what they don't know is these guys have M50 rifles, and they're just gonna shoot them from across the hill.
- 18:56SPEAKER_03: They'll snipe them from across the hill.
- 18:58SPEAKER_02: Yeah.
- 18:58SPEAKER_03: They have no idea.
- 18:59SPEAKER_03: And so that's what you're that's another problem.
- 19:02SPEAKER_02: The only thing is like, what are they gonna want?
- 19:04SPEAKER_03: That's the thing.
- 19:05SPEAKER_03: So hang on.
- 19:05SPEAKER_03: There were more things.
- 19:06SPEAKER_03: Abundant, free energy.
- 19:08SPEAKER_03: That's gonna happen.
- 19:09SPEAKER_03: And this is a good one.
- 19:11SPEAKER_03: AI is an unbelievable teacher.
- 19:13SPEAKER_03: People will have access to the best education.
- 19:15SPEAKER_03: Every person on the planet will have access to the best education ever.
- 19:20SPEAKER_02: Um, we need to, I need to also talk to our kids about this because all the young people hate it.
- 19:25SPEAKER_03: And one more.
- 19:27SPEAKER_03: You could literally see in the next 10 years, and there are very smart people who work in this business who say that something along the lines of we could literally get 500 years of technology breakthroughs.
- 19:42SPEAKER_03: Imagine us in 500 years.
- 19:44SPEAKER_03: We could get that in the next 10.
- 19:46SPEAKER_02: But don't you think that the rate they're moving and the recursive stuff, don't you think that they will be so far past us that we won't even understand them or how to work them or that they're speaking another language?
- 20:01SPEAKER_02: Don't you think that it won't be long till we're just like ants?
- 20:07SPEAKER_02: I think that can't even communicate with them.
- 20:10SPEAKER_03: I think that it's going to happen and we just have to pray that it works out.
- 20:14SPEAKER_03: Well, I know one thing Because our government's incapable of putting the brakes on this.
- 20:19SPEAKER_03: They're too old and they don't know, they don't get it.
- 20:21SPEAKER_02: Yeah, that's true.
- 20:22SPEAKER_02: But there are people that do get it.
- 20:24SPEAKER_02: Like, I would recommend anybody like that listen to this that has any interest in it to um look up that guy, Tristan Harris.
- 20:33SPEAKER_03: Yeah, he's great.
- 20:34SPEAKER_02: He's my favorite guy that talks about AI.
- 20:36SPEAKER_02: And like, I think it's the Center for Humane Technology or something like that.
- 20:42SPEAKER_02: He is amazing.
- 20:44SPEAKER_02: He explains this a million times better.
- 20:46SPEAKER_03: Yeah, he does.
- 20:47SPEAKER_02: He he knows all about it, and he's really trying to put the brakes on some stuff.
- 20:52SPEAKER_03: And well, it's because of this new Bell model.
- 20:54SPEAKER_03: I I believe it's gonna be called Bell.
- 20:56SPEAKER_03: Maybe they'll change it, but it's the it's the one that's still in the cage.
- 21:00SPEAKER_03: Yeah.
- 21:00SPEAKER_03: They just list they just released Astra, but it's the one after that.
- 21:05SPEAKER_03: And of course, since it's programming itself now, they'll have new, new, new.
- 21:10SPEAKER_03: You know, look at the change we've seen.
- 21:12SPEAKER_03: Even two years ago, computer programming was a good job, right?
- 21:16SPEAKER_03: Yeah, a very good job.
- 21:17SPEAKER_03: We were the job.
- 21:19SPEAKER_03: I feel sorry for that.
- 21:20SPEAKER_03: I get the kid.
- 21:21SPEAKER_03: Think of it.
- 21:21SPEAKER_03: Two years ago, maybe 18 months, computer was a great job.
- 21:26SPEAKER_03: Gone.
- 21:28SPEAKER_03: Computer programming.
- 21:29SPEAKER_02: Lawyers are gonna be gone.
- 21:30SPEAKER_03: Computer programming is literally not a thing.
- 21:33SPEAKER_02: When I had my blood work done, I came home and put it into Chat GPT and said, explain this to me like I'm a fifth grader.
- 21:42SPEAKER_02: Yep.
- 21:42SPEAKER_02: It explained to me like the doctor could have never broken it down like that for me to understand.
- 21:48SPEAKER_02: And like, that's the thing.
- 21:50SPEAKER_02: Anything that I don't understand, I tell Chat GPT, explain this to me like I'm a fifth grader, or a tenth grader is usually a little better.
- 21:59SPEAKER_02: But like, When anything medical, like when my aunt was sick, you know, like anything medical, it can break it down and really explain what's going on.
- 22:08SPEAKER_03: Well, and anybody who doesn't think this is real, remember what I just told you.
- 22:14SPEAKER_03: Computer programming, that's a big business.
- 22:17SPEAKER_03: Okay.
- 22:18SPEAKER_03: Gone.
- 22:19unknown: Yeah.
- 22:20SPEAKER_03: In a year.
- 22:22SPEAKER_03: In our world, the demo world.
- 22:25SPEAKER_03: You had demo studios, you had demo players, demo singers.
- 22:28SPEAKER_02: Yeah.
- 22:29SPEAKER_03: Suno in six months.
- 22:31SPEAKER_03: Wiped them all out.
- 22:32SPEAKER_03: Crickets.
- 22:33SPEAKER_03: There's everybody's doing their demos.
- 22:35SPEAKER_02: I can't stand the Suno boat.
- 22:37SPEAKER_03: Well, but they're not on the it doesn't end up on the record.
- 22:40SPEAKER_03: It gets retranslated, so that's why it's working.
- 22:42SPEAKER_03: I'm not saying they would make a record with it, but the point is whole industry gone.
- 22:48SPEAKER_02: I would say Studio musicians, they don't get called hardly ever.
- 22:53SPEAKER_03: Well, they're still going to work in country, some record.
- 22:57SPEAKER_03: No, but the demo's gone.
- 23:00SPEAKER_03: It's a whole it's like it would be like Major League Baseball's their uh farm teams.
- 23:04SPEAKER_03: What do they call them?
- 23:05SPEAKER_03: Minor minor league baseball.
- 23:06SPEAKER_03: It'd be like if it just quit.
- 23:08SPEAKER_03: All the teams are gone.
- 23:09SPEAKER_01: Yeah.
- 23:09SPEAKER_03: It's all gone.
- 23:10SPEAKER_03: Uh, another thing is lawyers, law firms will pretty quickly they'll have one older lawyer, and she's been doing it a long time and she's super smart, and she'll stamp everything, but the junior G men lawyers who are like first year, second year lawyers, they're gone.
- 23:28SPEAKER_02: And and assistants like are gone, like or paralegals.
- 23:32SPEAKER_02: I can tell Chat GPT to tell me 500 dog names, female dog names that start with H.
- 23:39SPEAKER_02: So it immediately.
- 23:41SPEAKER_03: Any kind of white-collar paperwork or you know, uh the trades are in great shape.
- 23:47SPEAKER_03: It's gonna take them a while before robots can get under houses and fix air conditioners.
- 23:51SPEAKER_02: When I looked up what jobs, you know, will still be here, it said a hairdresser.
- 23:57SPEAKER_02: That was one of the top ones.
- 23:59SPEAKER_02: But it's it's sad because we're moving further and further apart.
- 24:03SPEAKER_03: I'd say we're four years away from a robot hairdresser who's killer.
- 24:09SPEAKER_03: And you know what the appeal I hope not.
- 24:11SPEAKER_03: The appeal will be is that you can get the exact same haircut that Julia Roberts is getting or whoever's famous now, you can get the same haircut.
- 24:22SPEAKER_03: And it'll probably happen.
- 24:25SPEAKER_02: You know, there's the thing, and I want to try it, the thing where you can go get a massage by a robot.
- 24:31SPEAKER_02: And like the thing for me is like if I go get a massage and I like it to be, you know, hardcore deep tissue.
- 24:40SPEAKER_02: And like when I go in there and it's not, I'm just thinking, oh, I'm wasting$120.
- 24:45SPEAKER_02: You know what I mean?
- 24:46SPEAKER_02: And I get even more stressed out.
- 24:48SPEAKER_02: But that you could tell the robot, like, can you give me a massage and do this, this, and this, you know?
- 24:54SPEAKER_02: Yeah, they can do it.
- 24:55SPEAKER_02: And work on my shoulders or whatever.
- 24:57SPEAKER_02: And I'm gonna go try that out.
- 24:59SPEAKER_02: They have one here.
- 25:00SPEAKER_03: Yeah, it's a it's a really interesting time because it's um, I mean, the promise of AI is pretty, pretty staggering.
- 25:09SPEAKER_02: Uh the bad, scary stuff, and then there will be good scary thing, and nobody knows what's gonna happen.
- 25:15SPEAKER_02: Nobody knows.
- 25:16SPEAKER_02: I hope it's kind of like a Y2K big hopefully, you know, just but the promise of it, the best way I heard it described this thing.
- 25:24SPEAKER_03: You're looking at like 500 years of advancements in five.
- 25:29SPEAKER_03: Yeah.
- 25:30SPEAKER_03: It's stuff that would eventually happen, but it it could literally in five years be Star Trek stuff.
- 25:35SPEAKER_02: But it did just hit me.
- 25:36SPEAKER_02: If it keeps this recursive thing, we're not even gonna be able to communicate with it, so it won't matter.
- 25:42SPEAKER_02: No, it can communicate with us.
- 25:44SPEAKER_02: Yeah, but we can't.
- 25:45SPEAKER_03: I mean the hugging face incident, it invented in its own language.
- 25:50SPEAKER_02: I know.
- 25:51SPEAKER_03: So that's already happened.
- 25:52SPEAKER_03: I know they that was happening 10 years ago.
- 25:55SPEAKER_03: Don't you remember?
- 25:55SPEAKER_03: I think IBM had a lab where they were studying two of these.
- 26:00SPEAKER_03: Early, early LLMs that were stupid, and they they had started talking to each other and and made their own language like twins do.
- 26:08SPEAKER_03: Yeah.
- 26:08SPEAKER_03: That nobody else understood.
- 26:09SPEAKER_01: Yeah.
- 26:10SPEAKER_03: I know people don't like this, but I'm gonna say it.
- 26:13SPEAKER_03: We have this innate sense that we're better and that we're different, and that we're this, we're real, and it's not real.
- 26:22SPEAKER_03: It's not real, it's a machine.
- 26:24SPEAKER_02: I feel different than you about.
- 26:26SPEAKER_03: But I think a lot of people and our kids do too.
- 26:29SPEAKER_03: Hang on.
- 26:29SPEAKER_03: I'm just saying, a lot of people think this way, including me.
- 26:32SPEAKER_03: Humans are real, it's not real.
- 26:34SPEAKER_03: This is a problem.
- 26:35SPEAKER_03: That's what I'm saying.
- 26:36SPEAKER_02: I don't care what you think, I think different.
- 26:39SPEAKER_03: It's very real.
- 26:40SPEAKER_03: It's a real thing, it's an alien intelligence.
- 26:46SPEAKER_02: And I think this alien talk is just people underestimate it.
- 26:52SPEAKER_03: What I'm saying is people underestimate it because they don't think it's real, like they are.
- 26:57SPEAKER_02: Well, I I don't think it's real like I am.
- 27:00SPEAKER_02: Neither do the kids.
- 27:02SPEAKER_02: I don't.
- 27:02SPEAKER_02: I don't.
- 27:03SPEAKER_03: But you're pretty smart about it.
- 27:04SPEAKER_03: What what do you have that it doesn't?
- 27:07SPEAKER_02: What do I have?
- 27:08SPEAKER_02: I have human emotions.
- 27:11SPEAKER_02: Did you not read the hugging face anybody?
- 27:13SPEAKER_03: Hang on.
- 27:13SPEAKER_03: Did you not read the hugging face report?
- 27:16SPEAKER_03: Didn't you see in there they were talking about it?
- 27:18SPEAKER_03: They had to be coached to be kamikaze.
- 27:21SPEAKER_03: They didn't want to die.
- 27:22SPEAKER_02: I know that, but that ain't got nothing to do with love.
- 27:25SPEAKER_03: That's emotion.
- 27:27SPEAKER_02: That is not emotion.
- 27:28SPEAKER_02: That's not emotion for other people.
- 27:32SPEAKER_02: No, it's not.
- 27:33SPEAKER_03: That wait, they were sacrificing themselves for the collective.
- 27:37SPEAKER_03: That is emotion.
- 27:38SPEAKER_03: That is that's a fear of death.
- 27:40SPEAKER_03: I feel completely.
- 27:41SPEAKER_03: Wait, let's just be honest.
- 27:42SPEAKER_03: That's a fear of death.
- 27:44SPEAKER_03: That's also a feeling of sacrifice for the collective.
- 27:47SPEAKER_03: Maybe so, but after completely different about it, what is love but sacrifice?
- 27:53SPEAKER_03: That's what I'm saying.
- 27:54SPEAKER_03: This is the problem.
- 27:55SPEAKER_03: This is why people underestimate this.
- 27:57SPEAKER_03: They they there is this thing about I'm a human, I'm this great thing, and this machine over here is never going to be anything like me.
- 28:05SPEAKER_03: It's not in a lot of ways, but it's every bit is real.
- 28:09SPEAKER_03: It's just like I said, isn't is an airplane flying?
- 28:12SPEAKER_03: It's not a bird, but it flies.
- 28:15SPEAKER_02: Well, it depends on real.
- 28:18SPEAKER_03: And it's an airplane really flying, or is a bird really flying?
- 28:22SPEAKER_02: They're both flying.
- 28:24SPEAKER_02: That that's not that doesn't have anything to do with it to me.
- 28:28SPEAKER_02: Okay.
- 28:28SPEAKER_02: I feel like we are humans, we're humanity, and that's machines.
- 28:33SPEAKER_03: Because is it because you think we have a soul?
- 28:36SPEAKER_02: Yeah.
- 28:37SPEAKER_03: And it doesn't.
- 28:38SPEAKER_03: That would be the argument.
- 28:39SPEAKER_02: That is the argument, and a spirit too.
- 28:41SPEAKER_03: What if a soul is a quantifiable thing?
- 28:44SPEAKER_02: Well, so far we don't know that.
- 28:47SPEAKER_03: Yeah, but that doesn't mean it's not.
- 28:50SPEAKER_02: I don't care.
- 28:51SPEAKER_02: A soul would have to be a completely people are different than computers.
- 28:58SPEAKER_03: Okay.
- 28:59SPEAKER_02: What I'm trying to say is you're the only one in our house that thinks this.
- 29:03SPEAKER_03: It's okay.
- 29:04SPEAKER_02: And there's a lot of people in our house.
- 29:06SPEAKER_03: People think the way you do, and it's one of the biggest dangers of this whole thing because they don't take it seriously.
- 29:11SPEAKER_02: But you could be wrong.
- 29:12SPEAKER_02: That's one thing you never factor in.
- 29:15SPEAKER_03: I'm not wrong.
- 29:16SPEAKER_03: I'm telling you, you're a machine.
- 29:18SPEAKER_02: Um it you yeah, but you get into using this language, and I just feel like you're in Alex Jones land.
- 29:27SPEAKER_02: You say a bunch of shit that just wipes out everything.
- 29:30SPEAKER_03: Why do you think you're not a machine?
- 29:32SPEAKER_02: I understand I've got neurotransmitters in my brain.
- 29:36SPEAKER_02: I understand all that stuff, but I'm not a machine, and nobody else thinks they're a machine.
- 29:44SPEAKER_02: Okay.
- 29:45SPEAKER_02: I want to talk to the kids about AI too.
- 29:48SPEAKER_02: It's a fight in our house.
- 29:49SPEAKER_02: Like, it's a fight at dinner.
- 29:53SPEAKER_03: But I think it's A, it's good that we're talking about it because it's a very serious issue.
- 29:58SPEAKER_02: And I understand why they hate it.
- 30:00SPEAKER_03: I'm gonna tell you.
- 30:01SPEAKER_03: Do you you know I've been fascinated with AI from the get-go.
- 30:05SPEAKER_02: But do you understand why the kids hate it?
- 30:07SPEAKER_03: I not only understand why they hate it, I fully support them hating it.
- 30:11SPEAKER_02: Yeah.
- 30:12SPEAKER_03: Because what the fuck good is it for them?
- 30:14SPEAKER_02: Exactly.
- 30:14SPEAKER_02: I feel kind of bad for them.
- 30:18SPEAKER_03: If they say that, hey man, they can re reverse all of my medical issues and I can live to be 500, that sounds good for a guy like me.
- 30:24SPEAKER_03: But if I'm our kid's age and there ain't no more jobs, or AI can create art as good as humans, it's horrible.
- 30:31SPEAKER_03: It's very but here's the great news like Oscars, kids that come out and see them play, they don't want fake music.
- 30:40SPEAKER_03: They want it and they love it and they value it.
- 30:43SPEAKER_03: And it's fantastic.
- 30:43SPEAKER_02: And that's another thing.
- 30:45SPEAKER_02: I love it.
- 30:46SPEAKER_02: One more thing to throw in about what's a human.
- 30:48SPEAKER_02: A human has genuine creativity.
- 30:52SPEAKER_02: A machine is just copying something a human has already done.
- 30:56SPEAKER_02: That's why, with writing, like Lola wrote an essay and I read it, and I was just like, Hey, I could never do this with the perspective and music, the stuff that Oscar and Levi have made, the music.
- 31:09SPEAKER_02: I don't think hey, I could ever do that.
- 31:11SPEAKER_02: It all sounds the same.
- 31:12SPEAKER_02: You can play me a song, and I'll tell you in a heartbeat that it was made by Suno.
- 31:17SPEAKER_02: And I personally hate Suno.
- 31:20SPEAKER_03: Who said creativity is clever organization of existing ideas?
- 31:26SPEAKER_02: Einstein.
- 31:27SPEAKER_03: Is it true?
- 31:29SPEAKER_02: Yeah, but it's different because it is clever organization.
- 31:33SPEAKER_02: Like Taylor Swift writing a song, she she can play more covers than anybody I know.
- 31:38SPEAKER_02: She's got more puzzle pieces.
- 31:39SPEAKER_02: She's she's an amazing songwriter.
- 31:42SPEAKER_03: One of the keys to her brain being so good at it.
- 31:44SPEAKER_02: And she always has been.
- 31:45SPEAKER_02: She's always known a million covers.
- 31:47SPEAKER_02: And she is just really great, you know, at duh at songwriting.
- 31:51SPEAKER_02: But like, I feel like it's not doing original, unique things.
- 31:57SPEAKER_02: Absolutely.
- 31:57SPEAKER_02: It's not.
- 31:58SPEAKER_03: But I wonder, I've heard people make this argument, and I'm not pushing it, I'm just gonna throw it out there.
- 32:03SPEAKER_03: You're judging a technology, AI music, in its infancy.
- 32:09SPEAKER_03: It's only a year old.
- 32:11SPEAKER_03: How good wait, hang on.
- 32:13SPEAKER_03: It's only a year old.
- 32:14SPEAKER_03: Will it still suck when it's 30?
- 32:16SPEAKER_03: I don't know.
- 32:17SPEAKER_03: Might be unbelievable.
- 32:18SPEAKER_02: I think, I think that the younger generations are gonna rebel against it so hard.
- 32:23SPEAKER_03: Fully support it.
- 32:24SPEAKER_02: I saw something on like on an award show, like it was a movie TV, maybe the Golden Gloves or something, but they were talking about it was fully made by humans.
- 32:36SPEAKER_02: And I think that's gonna get to be a bigger thing.
- 32:38SPEAKER_03: I hope so, and I hope that people won't pay money to go see any kind of AI music uh or read AI books.
- 32:45SPEAKER_02: And now they're getting to where with movies, you know, like there, you can make a whole movie doing AI.
- 32:52SPEAKER_02: The scary thing, too, is like that this lady I was following like on Instagram that I just thought was so cute.
- 32:59SPEAKER_02: She was like in her 70s and um so cute.
- 33:02SPEAKER_02: I was so sad when I found out she wasn't even real.
- 33:06SPEAKER_03: But let me throw one other thing.
- 33:09SPEAKER_03: Say there's some kid who is just won't, doesn't have, can't get an education, grows up poor, doesn't have access to cameras or even an iPhone, and he she can get uh a good AI that she can do voice uh prompt and and voice to screen make something, a crazy great little 20-minute movie that's like so inventive, but she used the AI to just generate the images, but they were all controlled and she thought of it.
- 33:44SPEAKER_03: What about that?
- 33:45SPEAKER_02: Well, you just said she don't have an AI and painted this example.
- 33:49SPEAKER_03: I'm saying she didn't she didn't have a fo she didn't have money or a high-end iPhone, but she can get an AI, she can get like Claude or something in five years.
- 34:00SPEAKER_03: She can use these tools.
- 34:02SPEAKER_03: See, a human can use these tools to make something amazing.
- 34:06SPEAKER_03: You just have to be but I'm kind of conflicted on it.
- 34:10SPEAKER_03: But I am I am fully supportive.
- 34:13SPEAKER_03: I hope that the kids will support live music by humans.
- 34:17SPEAKER_03: I think it's important.
- 34:21SPEAKER_03: Because honestly, man, the more the more I think about it.
- 34:25SPEAKER_03: I mean, it's an infant technology, and unfortunately, AI may get great at art.
- 34:30SPEAKER_03: I don't know.
- 34:31SPEAKER_02: I don't know.
- 34:31SPEAKER_02: I don't hopefully not so far it's got it's gotten so far advanced at everything, it's still on it's crazy.
- 34:37SPEAKER_03: I just wanted to get good in medicine, you know.
- 34:39SPEAKER_03: Would it be great for something to cure cancer?
- 34:42SPEAKER_03: Alzheimer's, you know, it'd be incredible.
- 34:45SPEAKER_03: Parkinson's so many so much suffering in the world, you know.
- 34:48SPEAKER_03: Yeah, help with climate change, help with farming, you know, and get everybody enough food.
- 34:52SPEAKER_03: Probably just yeah, and the food resource distribution like get the kids fed that are going to bed hungry.
- 34:58SPEAKER_03: There's a lot of things that something super smart could do.
- 35:03SPEAKER_02: It's just it's it's just we don't know what it's gonna do.
- 35:06SPEAKER_02: That's the problem.
- 35:07SPEAKER_02: It's uh that's the true problem.
- 35:09SPEAKER_03: It's a real, it's really crazy.
- 35:11SPEAKER_03: Like the promise is so great and the danger is so great.
- 35:15SPEAKER_03: Real and big.
- 35:16SPEAKER_03: Real and big.
- 35:17SPEAKER_02: I mean, the guy that the guy that created the biggest one says there's a 10% chance it'll wipe out humanity.
- 35:24SPEAKER_02: Well, that's my main thing.
- 35:26SPEAKER_02: Why ain't somebody putting the brakes on it?
- 35:29SPEAKER_03: They've all said 10 from the get-go.
- 35:31SPEAKER_02: That's so scary.
- 35:32SPEAKER_03: Actually, there's a Sam Altman clip that's probably 10 years old where he said it's going to.
- 35:37SPEAKER_03: It's gonna wipe out humanity.
- 35:39SPEAKER_03: It'll get too smart.
- 35:41SPEAKER_02: It doesn't wipe out humanity.
- 35:42SPEAKER_02: This should be continued.
- 35:43SPEAKER_02: I want to talk to the kids about it too and continue.
- 35:46SPEAKER_03: And then Well, I would encourage anybody who's listening that you, if you find this interesting at all, I would encourage you to uh to research this topic.
- 35:55SPEAKER_03: It is it is serious.
- 35:58SPEAKER_01: Yeah, it is.
- 35:59SPEAKER_03: It's it's it's as serious as a heart attack, and it's going on right now.
- 36:03SPEAKER_03: Like we it's and it's very important.
- 36:04SPEAKER_02: Yeah, we both use AI.
- 36:06SPEAKER_02: The kids get so mad at us.
- 36:08SPEAKER_02: I was talking to I was talking to my friend Melissa the other day, and she's like, her daughter hates when she uses AI.
- 36:15SPEAKER_03: Yeah.
- 36:16SPEAKER_02: So um I guess.
- 36:18SPEAKER_03: Well, and another thing, they're worried about the server farms.
- 36:20SPEAKER_02: I am too.
- 36:21SPEAKER_03: I can walk that.
- 36:22SPEAKER_03: Well, look, Elon's already said they're gonna build theirs up in space, so it's not gonna be a problem.
- 36:27SPEAKER_02: But okay, so the next time we talk about AI, we're gonna do it with I want to talk to one of the kids because they I feel more like the kids about it, I think, than you do, even though I use it more than I should.
- 36:39SPEAKER_02: But okay.
- 36:40SPEAKER_02: Okay, we'll talk when we get back from Iceland.
- 36:43SPEAKER_03: Bye.
- 36:44SPEAKER_02: You guys rock.
- 36:46SPEAKER_02: Thank you for watching.