Latest / Breaking Points with Krystal and Saagar / 10/15/26: Kamala Weak Biden Criticism To Tee Up 2028, AI Researchers Smoke Expecting AI To Cure Cancer

Transcript

10/15/26: Kamala Weak Biden Criticism To Tee Up 2028, AI Researchers Smoke Expecting AI To Cure Cancer · Breaking Points with Krystal and Saagar

• 0:00
  Speaker 1: Had a busy day, don't have time to listen to
• 0:01
  Speaker 1: the show?
• 0:02
  Speaker 2: Catch up on Breaking Points with our free daily newsletter.
• 0:05
  Speaker 1: You get our breakdown of every daily story.
• 0:08
  Speaker 2: And a lot of links in your inbox every day.
• 0:11
  Speaker 3: Head over, breakingpoints.com. You can sign up for the newsletter
• 0:14
  Speaker 3: at the top for free today.
• 0:19
  Speaker 2: Kamala Harris admitting really for the first time mistakes in
• 0:22
  Speaker 2: the Biden policy vis-a-vis Gaza as she sets up a
• 0:25
  Speaker 2: potential 2020 run.
• 0:26
  Speaker 4: Let's take a listen.
• 0:27
  Speaker 5: America must use our leverage... to ensure our actions align
• 0:33
  Speaker 5: with our interests. It is in our interest to hold
• 0:37
  Speaker 5: our allies accountable for their share of global defense, to
• 0:41
  Speaker 5: ensure trade benefits, American workers and consumers, and to uphold
• 0:47
  Speaker 5: our moral obligation to human rights and international law. To
• 0:54
  Speaker 5: be sure, America has at times fallen short on that one.
• 1:00
  Speaker 5: Most recently, the failure to use our leverage to curtail
• 1:05
  Speaker 5: the Israeli government's horrific conduct in the war in Gaza,
• 1:11
  Speaker 5: both under the current administration, which offers Netanyahu a blank check,
• 1:17
  Speaker 5: and our administration before, where we should have acted sooner
• 1:23
  Speaker 5: and more forcefully.
• 1:25
  Speaker 2: Should have acted sooner and more forcefully, I would say,
• 1:29
  Speaker 2: a lot more than that.
• 1:30
  Speaker 1: I missed her sunny disposition.
• 1:32
  Speaker 2: Yeah, I mean, and this is what counts for like
• 1:35
  Speaker 2: a big break with Biden, you know, a shocking moment
• 1:39
  Speaker 2: from Kamala Harris because she's just so. First of all,
• 1:42
  Speaker 2: devoid of any actual compass or principles. She's just constantly
• 1:46
  Speaker 2: buffeted by where she thinks the political wins are. And
• 1:49
  Speaker 2: second of all, just so cautious. And this really doomed
• 1:52
  Speaker 2: her campaign, ultimately, that she would not break with Biden,
• 1:55
  Speaker 2: in particular on this issue where so many young people
• 1:58
  Speaker 2: and so much of the base of the party.
• 2:00
  Speaker 4: Was completely disgusted.
• 2:02
  Speaker 2: And now you have Tim Walz coming out and saying, hey,
• 2:04
  Speaker 2: I think this was a problem for us in terms
• 2:05
  Speaker 2: of the election.
• 2:06
  Speaker 1: Yeah, no shit, Tim. But it's like, how long did
• 2:08
  Speaker 1: it take?
• 2:09
  Speaker 3: You know what's fascinating to me is, because we just
• 2:11
  Speaker 3: did these segments about the candidate quality, compare her to
• 2:15
  Speaker 3: any of these upstarts.
• 2:16
  Speaker 1: Tallarico, Ossoff. I mean, I could go down the list.
• 2:20
  Speaker 4: The guy in Mississippi.
• 2:23
  Speaker 2: Angie Nixon is such a badass.
• 2:27
  Speaker 3: Any of these people. And I'm not talking about politics.
• 2:30
  Speaker 3: I'm saying purely communication. I'm like, oh, my God, I
• 2:33
  Speaker 3: forgot how awful she was. Like this woman cannot speak properly.
• 2:37
  Speaker 3: And everything is just down the middle. Like everything is planned.
• 2:41
  Speaker 4: Carefully crafted.
• 2:42
  Speaker 1: You believe in nothing.
• 2:43
  Speaker 4: Scripted.
• 2:44
  Speaker 6: Yeah.
• 2:44
  Speaker 3: And yet I think she's running. Let's put the next
• 2:47
  Speaker 3: one up here. You know, she's barnstorming. Here we have
• 2:50
  Speaker 3: the latest news. Kamala headed to Nevada, Mississippi, New Hampshire, Wisconsin.
• 2:55
  Speaker 1: And Georgia.
• 2:56
  Speaker 3: I know Abdul took a lot of shit for doing
• 2:58
  Speaker 3: an event with her, which, look, he was in an
• 3:00
  Speaker 3: impossible position.
• 3:01
  Speaker 1: Like, obviously, you know, you can't turn down Kamala Harris.
• 3:04
  Speaker 7: Yeah.
• 3:05
  Speaker 3: But I just don't see how she could be a
• 3:07
  Speaker 3: net positive to your campaign. Maybe in the very specific
• 3:11
  Speaker 3: case of Michigan.
• 3:12
  Speaker 7: Yes.
• 3:12
  Speaker 3: Because you need black voters where he's struggling with in Detroit,
• 3:15
  Speaker 3: et cetera, although, you know, she did lose Michigan.
• 3:17
  Speaker 1: But Nevada?
• 3:18
  Speaker 8: Yeah.
• 3:19
  Speaker 3: Mississippi, New Hampshire, Wisconsin, Georgia. So I can think of Mississippi, Wisconsin, Georgia.
• 3:25
  Speaker 3: Those all have important black constituencies. Although I don't think,
• 3:28
  Speaker 3: you know, John Ossoff exactly needs her help winning election.
• 3:31
  Speaker 2: But Keisha Lance Bottoms does.
• 3:33
  Speaker 3: True.
• 3:33
  Speaker 7: That's a good point. I hadn't thought about that.
• 3:35
  Speaker 3: But, you know, Nevada and New Hampshire, I'm like, no,
• 3:38
  Speaker 3: you should keep this lady away. She's not somebody who
• 3:41
  Speaker 3: you want on your side. But nonetheless, the groundwork is
• 3:43
  Speaker 3: being laid. The Israel stuff, obviously she's trying to do more.
• 3:46
  Speaker 3: She's hitting the campaign trail. She had her book, you know,
• 3:50
  Speaker 3: 107 Days or whatever, basically said nothing of genuine consequence.
• 3:55
  Speaker 3: And then we're about to play this Gavin thing, but
• 3:58
  Speaker 3: I don't know. I mean, all signs indicate I think
• 4:00
  Speaker 3: she's going to run. Like, I think it's going to happen.
• 4:03
  Speaker 1: I maybe.
• 4:04
  Speaker 2: I mean, certainly the campaign travel would indicate it. But
• 4:07
  Speaker 2: I'm also looking at this and I'm like, where are
• 4:09
  Speaker 2: you been for the past two years? You know, I
• 4:11
  Speaker 2: think that if she does run, voters are going to
• 4:14
  Speaker 2: see it feel the same way. I mean, on the
• 4:16
  Speaker 2: ballot with Doge and with the ice invasions and the
• 4:19
  Speaker 2: National Guard deployments and the drug boats and the Gaza
• 4:23
  Speaker 2: and Iran war.
• 4:24
  Speaker 4: And she's just been absent from all of this.
• 4:27
  Speaker 2: And to your point, like as much as there are
• 4:30
  Speaker 2: plenty of segments of the American population that are like
• 4:33
  Speaker 2: whatever about Kamala Harris, if you look at the polling,
• 4:37
  Speaker 2: you know, black voters really do still admire her and
• 4:40
  Speaker 2: really do look first female black vice president. Very important,
• 4:44
  Speaker 2: you know, trailblazing role. And there is a genuine like
• 4:46
  Speaker 2: affection and affinity and admiration there.
• 4:49
  Speaker 4: And you didn't.
• 4:50
  Speaker 2: Spend any time getting your hands dirty, taking any risks
• 4:54
  Speaker 2: during a time period when, you know, the Democratic base
• 4:57
  Speaker 2: felt this was very existential. And so even now, you
• 5:01
  Speaker 2: can hear in her comments how constrained she.
• 5:04
  Speaker 1: Is, right?
• 5:04
  Speaker 2: It's a Gaza war. She'll never say it's a genocide
• 5:07
  Speaker 2: because she can't, because then that would admit her complicity
• 5:10
  Speaker 2: in genocide as part of the Biden-Harris administration.
• 5:13
  Speaker 4: But that's also just how she operates.
• 5:15
  Speaker 2: So safe, so cautious, which is why You know, Kyle
• 5:18
  Speaker 2: and I actually have a disagreement about this because when
• 5:20
  Speaker 2: you poll the Democratic primary electorate of who they are
• 5:23
  Speaker 2: interested in 2028, she's still quite high.
• 5:24
  Speaker 5: She's high.
• 5:27
  Speaker 2: I just look at these clips of her and I go,
• 5:29
  Speaker 2: there's no way that even if she runs, that she's
• 5:32
  Speaker 2: going to be a front runner. Plus, why would you
• 5:34
  Speaker 2: go with someone who was already a proven loser? That
• 5:37
  Speaker 2: makes absolutely no sense to me. And if there's one
• 5:39
  Speaker 2: thing about Democrats, one, they definitely want to win. As
• 5:42
  Speaker 2: for whether or not she's going to run, if I
• 5:44
  Speaker 2: was just looking at these things, I would agree with
• 5:46
  Speaker 2: you of like, yeah, she's laying the groundwork. She's doing
• 5:49
  Speaker 2: so inflexibly. She hasn't done enough, but that's just kind
• 5:52
  Speaker 2: of her style. She is actually planning to run. However,
• 5:55
  Speaker 2: the thing that does give me pause is, is this
• 5:57
  Speaker 2: thing that Gavin Newsom said to Jake Tapper in this
• 6:00
  Speaker 2: bizarre fly-fishing moment they were having. Let's go ahead and
• 6:03
  Speaker 2: take a listen to that.
• 6:03
  Speaker 7: It's going to be a lot of people.
• 6:05
  Speaker 2: Yeah.
• 6:05
  Speaker 7: It's a lot of Democrats. 31 flavors. You're not even
• 6:08
  Speaker 7: the only Californian. No, right. You've got Ro Khanna, Kamala.
• 6:13
  Speaker 3: Yeah.
• 6:13
  Speaker 7: I don't know if she runs, but we'll see. I
• 6:15
  Speaker 7: know she's called around. Yeah. But I wouldn't run if
• 6:18
  Speaker 7: she ran. If she runs, you will not?
• 6:21
  Speaker 9: Of course not.
• 6:22
  Speaker 7: Why would I? I wouldn't do that to her. I
• 6:25
  Speaker 7: wouldn't do it to me. Why would I want to
• 6:26
  Speaker 7: do that to you? I wouldn't waste everyone's time. Who
• 6:28
  Speaker 7: needs that?
• 6:30
  Speaker 2: So he says quite definitively there without being asked or prompted,
• 6:33
  Speaker 2: if she runs, I'm not going to run. Like, he
• 6:35
  Speaker 2: clearly wanted to get that out there and probably curry
• 6:38
  Speaker 2: favor with her and, you know, with.
• 6:41
  Speaker 4: Whoever her supporters are.
• 6:42
  Speaker 2: But it made me think that he must know she's—because
• 6:45
  Speaker 2: he obviously is desperate to be president. It made me
• 6:48
  Speaker 2: think that he must know something behind the scenes that
• 6:50
  Speaker 2: she's probably not really going to run. That's what that
• 6:52
  Speaker 2: indicated to me, that this was like a freebie for him.
• 6:55
  Speaker 2: He gets a posture like, oh, he's the good guy.
• 6:56
  Speaker 2: He'd step aside for— Kamala Harris, but in reality has
• 7:00
  Speaker 2: some inside info that she's probably not going to run anyway.
• 7:02
  Speaker 4: But maybe I'm reading too much into this. I don't know.
• 7:04
  Speaker 1: Maybe you are.
• 7:05
  Speaker 3: I just something in my gut says she's going to run. And,
• 7:09
  Speaker 3: you know, take a look at the map. Right. Because remember,
• 7:12
  Speaker 3: we're not starting in Iowa. We're not starting in New Hampshire.
• 7:15
  Speaker 3: We're starting in South Carolina, which is a huge advantage
• 7:19
  Speaker 3: to the establishment Democratic Party and specifically to the remember,
• 7:23
  Speaker 3: who's the DNC chair who rigged this entire system for
• 7:26
  Speaker 3: Joe Biden, for Kamala And these types of people in
• 7:28
  Speaker 3: the beginning, Jamie Harrison, the failed idiot, right, who refuses
• 7:32
  Speaker 3: to even publish their own autopsy, which blames Biden, doesn't
• 7:35
  Speaker 3: mention Biden's age, doesn't mention Israel.
• 7:37
  Speaker 7: I mean, it's a joke.
• 7:38
  Speaker 1: And so this is all the groundwork is there.
• 7:41
  Speaker 3: This exact constituency, she wins one state and it sets
• 7:44
  Speaker 3: her up for some sort of Super Tuesday type sweep.
• 7:47
  Speaker 1: It's exact.
• 7:48
  Speaker 3: And then Demer, like, oh, well, maybe she is a
• 7:49
  Speaker 3: winner this time around. I just I don't know. She
• 7:53
  Speaker 3: has a significant amount of affection. A lot of people
• 7:55
  Speaker 3: still have common envy, like rose-colored glasses, like, well, she
• 7:58
  Speaker 3: would have been so much better than Trump.
• 7:59
  Speaker 1: Like, you could see it. And so, I don't know.
• 8:02
  Speaker 1: I mean, you bring on the right candidate this time around.
• 8:04
  Speaker 4: She'd have to get through debates, though.
• 8:06
  Speaker 1: Yeah, I agree. But, I mean, look, I'm with you.
• 8:10
  Speaker 3: I think she's one of the worst candidates in modern
• 8:12
  Speaker 3: American history. I mean, obviously, I think she'd be a
• 8:15
  Speaker 3: terrible president. However, you know, all of that being said, like,
• 8:19
  Speaker 3: she's got the money behind her. She's got the right people.
• 8:23
  Speaker 3: One other sign, I don't know if you saw this,
• 8:25
  Speaker 3: is that her husband has been going around, according to
• 8:27
  Speaker 3: the Wall Street Journal, telling everyone he hopes that she
• 8:29
  Speaker 3: doesn't run, which makes me be like, huh, so hope
• 8:32
  Speaker 3: you don't run. Not she's not going to run, as
• 8:36
  Speaker 3: in he's like trying to cajole her into not running.
• 8:40
  Speaker 1: So it's kind of, look, we're T. Lee reading.
• 8:42
  Speaker 4: I don't know about that one.
• 8:44
  Speaker 3: Yeah.
• 8:44
  Speaker 2: I would almost read that as a counter indicator because
• 8:46
  Speaker 2: that would give her, no, because he wouldn't, he's not
• 8:49
  Speaker 2: running to the wall street journal or whatever to like
• 8:52
  Speaker 2: indirectly talk to his wife. That would be them putting
• 8:56
  Speaker 2: out a rationale for why she's going to take a path.
• 8:59
  Speaker 3: Well, he doesn't want her to run because it would
• 9:00
  Speaker 3: disrupt his own business interests. So it's not for like
• 9:03
  Speaker 3: any particular, um, for, you know, It's not really for
• 9:06
  Speaker 3: benevolent purposes.
• 9:08
  Speaker 2: I mean, I'm sure it was very hard on them.
• 9:09
  Speaker 2: I have no doubt about that in any number of ways. Yeah, interpersonal,
• 9:14
  Speaker 2: all that. But there is another California drama going on
• 9:18
  Speaker 2: here that we wanted to highlight between specifically Ro Khanna
• 9:21
  Speaker 2: and Gavin Newsom. We can put D4 up on the screen.
• 9:26
  Speaker 2: California has on the ballot this billionaire tax, which has
• 9:31
  Speaker 2: become a major issue. And there's a lot of complicated
• 9:34
  Speaker 2: political dynamics behind this thing, even among progressives, because one
• 9:39
  Speaker 2: of the guys who's championing it is very polarizing this
• 9:42
  Speaker 2: labor figure.
• 9:43
  Speaker 4: However, Roe, to his credit, representing.
• 9:46
  Speaker 2: Silicon Valley, has been one of the foremost proponents of
• 9:50
  Speaker 2: passing this billionaire wealth tax on the ballot. And Gavin
• 9:54
  Speaker 2: has positioned himself, unsurprisingly, on the other side. However, it's
• 9:58
  Speaker 2: not just that Gavin is like... saying that he's going
• 10:01
  Speaker 2: to vote against it. He has really allowed himself and
• 10:04
  Speaker 2: made himself the sort of poster child of arguing against
• 10:08
  Speaker 2: it and has positioned himself very clearly on the side
• 10:11
  Speaker 2: of the billionaires. And it's not a surprise to me
• 10:13
  Speaker 2: because that's the way he's operated politically. That is his
• 10:16
  Speaker 2: source of power. And I do think very much that
• 10:19
  Speaker 2: if he runs in 2028, he will be the stalking
• 10:21
  Speaker 2: horse for the Silicon Valley elites because, you know, they
• 10:25
  Speaker 2: don't just play one side of the political aisle. They
• 10:27
  Speaker 2: got their guy in Trump, but they would if the
• 10:30
  Speaker 2: wheel turns and Democrats are ascending to power, they're going
• 10:32
  Speaker 2: to want their candidates on the other side. And Gavin, obviously,
• 10:34
  Speaker 2: is someone who would be very close to them, is
• 10:36
  Speaker 2: doing their bidding here on something.
• 10:37
  Speaker 4: That for whatever reason.
• 10:39
  Speaker 2: They see as existential and they're really, really committed to
• 10:42
  Speaker 2: opposing this ballot measure. So, you know, I think he
• 10:47
  Speaker 2: also is going to have some trouble, given where the
• 10:50
  Speaker 2: Democratic Party is today, with making himself such a prominent
• 10:54
  Speaker 2: opponent of taxing the wealth of billionaires.
• 10:57
  Speaker 3: I really just have no idea because it's about the landscape.
• 11:00
  Speaker 3: It's about his opponents. It also comes down to the
• 11:02
  Speaker 3: map about who's going to run. Ro Khanna, the point,
• 11:05
  Speaker 3: I mean, the thing is, is it's important. is you
• 11:08
  Speaker 3: have to have somebody on the early debate stage because
• 11:10
  Speaker 3: that's where some of these breakout moments can still happen
• 11:13
  Speaker 3: when you have like 16, 17 people.
• 11:15
  Speaker 1: On the stage.
• 11:16
  Speaker 3: Khanna or somebody like him would have to be on
• 11:18
  Speaker 3: the stage to call out Gavin. I actually don't trust
• 11:20
  Speaker 3: even the moderators. You're going to trust CNN moderator to
• 11:23
  Speaker 3: be like, why did you oppose the California billionaire?
• 11:25
  Speaker 4: What about that well-taxed Gavin?
• 11:27
  Speaker 1: They're not going to ask him.
• 11:28
  Speaker 3: And none of these reporters are either. You literally would
• 11:31
  Speaker 3: have to have a candidate to force his hand. The
• 11:33
  Speaker 3: other thing I pay attention to is.
• 11:35
  Speaker 2: But you know, even some of the more moderate, like
• 11:37
  Speaker 2: this is one thing I've been thinking about, the Josh Tureks,
• 11:40
  Speaker 2: the James Tallaricos. they have really incorporated this language about
• 11:44
  Speaker 2: the billionaire class into their campaigning, this populist language. Now,
• 11:49
  Speaker 2: they may not have all the policy positions of a
• 11:50
  Speaker 2: Bernie Sanders, but rhetorically, if you listen to them and
• 11:53
  Speaker 2: you listen to Bernie in 2016, they've adopted a lot
• 11:56
  Speaker 2: of his language. So I think there will be a
• 11:59
  Speaker 2: lot of people on the stage who would, whether it's
• 12:01
  Speaker 2: Roe or someone else, who would be itching to make
• 12:04
  Speaker 2: that case against Gavin Newsom.
• 12:05
  Speaker 1: That's a good point. That's true.
• 12:07
  Speaker 3: I mean, obviously, these people are all very ambitious. For example, D5,
• 12:11
  Speaker 3: here we have John Ossoff doing the classic, I pledge
• 12:14
  Speaker 3: to serve six years in the Senate.
• 12:16
  Speaker 7: Let's take a listen.
• 12:17
  Speaker 9: So will you pledge right now tonight to serve your
• 12:19
  Speaker 9: full six-year term if you win?
• 12:21
  Speaker 7: Yes, I will.
• 12:22
  Speaker 1: I have zero interest in running for president. I'm not
• 12:25
  Speaker 1: running for president. I'm running for reelection to the Senate.
• 12:29
  Speaker 1: I have zero interest in running for president. I'm running
• 12:31
  Speaker 1: for Senate.
• 12:32
  Speaker 3: Can I tell you about another senator who pledged to
• 12:34
  Speaker 3: full a serve six-year term?
• 12:36
  Speaker 1: His name was Barack Hussein Obama.
• 12:39
  Speaker 7: All right.
• 12:40
  Speaker 3: So just these things are meaningless. By the way, I
• 12:43
  Speaker 3: have no interest could also mean I have no current interest.
• 12:46
  Speaker 7: Yeah.
• 12:46
  Speaker 2: Or you could point to Susan Collins who pledged to
• 12:49
  Speaker 2: only serve two terms.
• 12:51
  Speaker 3: Yeah.
• 12:52
  Speaker 4: For however long.
• 12:53
  Speaker 7: Exactly.
• 12:53
  Speaker 1: It's a joke.
• 12:54
  Speaker 2: Didn't Joe Biden also, at the beginning of his career,
• 12:57
  Speaker 2: he said he'd.
• 12:58
  Speaker 4: Only served two terms or something like that?
• 12:59
  Speaker 3: Ends up serving for 99 years. You know, it's like,
• 13:03
  Speaker 3: it's a joke. So a lot of people took that
• 13:05
  Speaker 3: to the bank.
• 13:06
  Speaker 7: They're like, oh, he pledged.
• 13:07
  Speaker 1: I'm like, you haven't been around long enough.
• 13:09
  Speaker 7: Okay.
• 13:10
  Speaker 1: They all say it.
• 13:10
  Speaker 4: He's got to say that right now.
• 13:12
  Speaker 7: Exactly. Yeah.
• 13:13
  Speaker 1: Let's put D, by the way, Tallarico said the same thing.
• 13:15
  Speaker 3: I think if Tallarico wins, he would announce his candidacy
• 13:18
  Speaker 3: literally the next day, in my opinion.
• 13:20
  Speaker 1: Let's put D6 up here.
• 13:22
  Speaker 3: On the screen, Warnock also coming out and saying, my
• 13:25
  Speaker 3: prayer is I can be.
• 13:26
  Speaker 1: Used to do the most good with the gifts that
• 13:29
  Speaker 1: I have.
• 13:29
  Speaker 3: And he was, by the way, in New Hampshire when
• 13:32
  Speaker 3: he was asked about running for president. And he has
• 13:35
  Speaker 3: been saying since he was a boy for God to
• 13:37
  Speaker 3: direct his path. apparently, to the presidency.
• 13:40
  Speaker 4: And Warnock's another one, very talented.
• 13:42
  Speaker 2: He's been more under the radar with, you know, Ossoff
• 13:45
  Speaker 2: running for re-election and really taking the lead in the state.
• 13:48
  Speaker 2: But Warnock is also very talented and charismatic. I mean,
• 13:51
  Speaker 2: he's a preacher, you know.
• 13:52
  Speaker 1: He's a preacher.
• 13:53
  Speaker 3: What do we keep saying about black voters, how much
• 13:55
  Speaker 3: it matters, especially in the early states?
• 13:58
  Speaker 1: Exactly.
• 13:59
  Speaker 3: You are representative of a state which Trump won in 2024.
• 14:03
  Speaker 1: I think that's actually.
• 14:07
  Speaker 1: You know, you're talking about, you know, one of the
• 14:08
  Speaker 1: things with Democrats, why do they choose to vote Biden?
• 14:10
  Speaker 1: They want to win, right?
• 14:11
  Speaker 3: And so here, actually nominating this way, a Tallarico or
• 14:14
  Speaker 3: a Warnock or any of these people, people who represent
• 14:17
  Speaker 3: a state which Trump won, that's going to be like
• 14:19
  Speaker 3: massively important. Abdul would be certainly in that conversation. In
• 14:23
  Speaker 3: my opinion, to be a real credible contender, you know, to, again,
• 14:28
  Speaker 3: to just to defeat J.D., who would be the likely nominee.
• 14:32
  Speaker 3: You would have to say, I have a track record
• 14:35
  Speaker 3: of winning over these types of voters, and I'm not
• 14:38
  Speaker 3: going to abandon you on said policy, right? Yeah.
• 14:40
  Speaker 1: That's what a winner is going to look like. That's
• 14:42
  Speaker 1: projecting confidence.
• 14:44
  Speaker 3: Especially if you want to try and unite a coalition
• 14:46
  Speaker 3: of people who are dissatisfied with the Trump administration and
• 14:49
  Speaker 3: a traditional Democratic Party.
• 14:51
  Speaker 1: Like, that's exactly the type of person who you would need.
• 14:53
  Speaker 2: Yeah. And that would be a tough one for AOC
• 14:56
  Speaker 2: to navigate.
• 14:57
  Speaker 7: New York?
• 14:59
  Speaker 4: You want anything.
• 15:00
  Speaker 2: You know, very blue district in New York. And so
• 15:02
  Speaker 2: even though the positioning is where the Democratic base is,
• 15:06
  Speaker 2: they're always going to have that electability question. And that's
• 15:09
  Speaker 2: part of why these midterms are so incredibly important for
• 15:12
  Speaker 2: setting the future of the Democratic Party. Because if Abdul
• 15:16
  Speaker 2: is able to win in a state like Michigan, then
• 15:18
  Speaker 2: the left has more of an argument to make about, see,
• 15:20
  Speaker 2: we are electable. See, we can get elected even in
• 15:22
  Speaker 2: a swing state like Michigan. It's not just the commie
• 15:24
  Speaker 2: corridor in Brooklyn. But in terms of as an individual candidate,
• 15:30
  Speaker 2: the fact she doesn't have a track record of winning
• 15:32
  Speaker 2: in a difficult swing state, that's going to be something
• 15:35
  Speaker 2: she'll have to overcome.
• 15:37
  Speaker 7: No question.
• 15:38
  Speaker 3: Last thing, we had Buttigieg on Twitter. the airwaves in
• 15:42
  Speaker 3: a Fox News interview. He was also asked about Israel.
• 15:45
  Speaker 3: Let's take a listen.
• 15:46
  Speaker 7: Israel or Iran? You have 60 seconds.
• 15:48
  Speaker 6: And again, Rick, I'm just going to say that that's
• 15:49
  Speaker 6: a ridiculous hypothetical, because here's what's happening right now.
• 15:52
  Speaker 7: I'll ask one last time. Israel or Iran?
• 15:55
  Speaker 6: That's, again, a ridiculous question. I think both of them
• 15:58
  Speaker 6: have proven dangerous in their own way. One, at risk
• 16:01
  Speaker 6: of attacking us. One, getting us into wars. We have
• 16:03
  Speaker 6: no business fighting.
• 16:05
  Speaker 2: All right.
• 16:06
  Speaker 4: A lot of people looked at that and said, gee,
• 16:07
  Speaker 4: that should be a pretty easy answer. What would your
• 16:09
  Speaker 4: answer have been? No, Iran is our enemy.
• 16:14
  Speaker 7: Israel is not.
• 16:16
  Speaker 4: They're an ally.
• 16:17
  Speaker 7: So that's pretty clear to me.
• 16:20
  Speaker 1: Pete's still doing the Israel is our ally bit. So
• 16:22
  Speaker 1: I don't know.
• 16:23
  Speaker 2: I just when I see him, I'm like Timu John Ossoff.
• 16:26
  Speaker 4: Like you missed your moment.
• 16:27
  Speaker 2: There's more quality, talented, centrist corporate types on the table now.
• 16:32
  Speaker 1: Also, he served under Biden. He didn't say anything. I mean,
• 16:34
  Speaker 1: I'm curious. Do you think that will come in the debate?
• 16:36
  Speaker 1: It should, in my opinion. Someone needs to be like, hey,
• 16:39
  Speaker 1: you literally served on Joe Biden. You didn't say anything.
• 16:40
  Speaker 2: Of course.
• 16:41
  Speaker 1: Yeah, I mean, I hope someone says that.
• 16:43
  Speaker 4: Absolutely.
• 16:43
  Speaker 1: Who does like one of the, what was it, Castro?
• 16:46
  Speaker 1: I forget which Castro brother.
• 16:47
  Speaker 5: Julian, I think.
• 16:48
  Speaker 3: Right, who was demonized at the time. They're like, oh,
• 16:50
  Speaker 3: how dare you talk to our great states. But he
• 16:52
  Speaker 3: was obviously 100% correct about him being completely out to lunch.
• 16:57
  Speaker 3: I do hope that some of these people get held accountable.
• 16:59
  Speaker 2: Well, Pete has been asked a number of times about
• 17:01
  Speaker 2: Israel and every time he makes the complete mess of
• 17:03
  Speaker 2: it because, you know, Pete's masterful at like finding some
• 17:08
  Speaker 2: bobbing and weaving middle ground where he says a lot
• 17:11
  Speaker 2: of stuff but doesn't really say anything at all. And
• 17:13
  Speaker 2: you just can't do that with the Democratic base that
• 17:17
  Speaker 2: sees us as having been complicit in a genocide. You
• 17:19
  Speaker 2: are not going to find a middle ground there. I
• 17:21
  Speaker 2: don't care how much you poll tested and focus group
• 17:24
  Speaker 2: it and read your briefing book like you need to
• 17:27
  Speaker 2: take a stand. And so I think that is going
• 17:30
  Speaker 2: to be disqualifying for him. And not to mention, again,
• 17:32
  Speaker 2: I just think that the quality of the candidates that
• 17:35
  Speaker 2: have emerged who are much more direct, who are very impressive,
• 17:40
  Speaker 2: much more charismatic than he is, and will have had
• 17:44
  Speaker 2: a track record of winning in difficult states against something
• 17:46
  Speaker 2: that he can't really claim. Um, I think he's going
• 17:49
  Speaker 2: to be, you know, I think he's going to be
• 17:50
  Speaker 2: out in class.
• 17:51
  Speaker 3: I saw Tallarico handle this last night. He's like, yeah,
• 17:53
  Speaker 3: Israel has a right to defend itself. They don't have
• 17:54
  Speaker 3: a right to slaughter civilians. I mean, that's still, you know,
• 17:57
  Speaker 3: it's like kind of centrist, Adelson approved answer. All of
• 18:01
  Speaker 3: that said, it's still better said than missed, you know,
• 18:04
  Speaker 3: it's still, it's better communicated, uh, in this way. So sorry, Pete, um,
• 18:08
  Speaker 3: as usual winning mayor of South Bend, Indiana doesn't actually
• 18:11
  Speaker 3: prepare you for anything.
• 18:12
  Speaker 1: All right. We've got a great guest standing by Kevin Roos.
• 18:14
  Speaker 7: Let's get to it.
• 18:18
  Speaker 3: We're very excited now to be joined by Kevin Roos,
• 18:19
  Speaker 3: great friend of the show. He is the author of
• 18:22
  Speaker 3: a fantastic new book. Let's put it up there on
• 18:24
  Speaker 3: the screen. The AGI Chronicles, The Inside Story of the
• 18:27
  Speaker 3: Race to Create an Artificial Superintelligence. He's also the host
• 18:32
  Speaker 3: of a forthcoming podcast, The Machine Gods Pod. We will
• 18:35
  Speaker 3: have links in the description to both. So, Kevin, let's
• 18:38
  Speaker 3: talk a little bit about why you decided to write
• 18:40
  Speaker 3: the book. It's incredibly timely that it's coming out now.
• 18:42
  Speaker 3: I know that you obviously then have been at the
• 18:44
  Speaker 3: forefront of this for some two odd years now, getting
• 18:46
  Speaker 3: interested in the characters. Why did you decide to write it?
• 18:50
  Speaker 3: Why is it relevant in terms of today's discussion?
• 18:53
  Speaker 9: Yeah, I think the book is basically a capsule history
• 18:56
  Speaker 9: of the last 10 years in AI and this race
• 18:59
  Speaker 9: between open AI, anthropic, and Google DeepMind to build artificial
• 19:03
  Speaker 9: general intelligence. That's sort of the holy grail of computer science,
• 19:07
  Speaker 9: the AI that can do everything the human brain can do.
• 19:11
  Speaker 9: And I just started to feel at one point like
• 19:13
  Speaker 9: this story was slipping away. Like I was witnessing something
• 19:16
  Speaker 9: that was sort of the 21st century's Manhattan Project. And
• 19:19
  Speaker 9: like no one was writing down the details, the conversations,
• 19:22
  Speaker 9: the breakthroughs, the feuds, the rivalries. It was all very
• 19:26
  Speaker 9: interesting to me. And so I just thought people need
• 19:28
  Speaker 9: some history here to understand how we got to this
• 19:31
  Speaker 9: point where these three companies are racing to build something
• 19:34
  Speaker 9: that now they're telling us could kill us all. So
• 19:36
  Speaker 9: I think to understand that and to understand the mindset
• 19:39
  Speaker 9: of the people who are building this stuff You have
• 19:41
  Speaker 9: to go back in time to 2016, 2017, when they
• 19:45
  Speaker 9: were all working together at this little company called OpenAI.
• 19:49
  Speaker 7: You had Elon Musk.
• 19:50
  Speaker 9: And Sam Altman and Ilya Satskovor and Dario Amede were
• 19:53
  Speaker 9: colleagues then. And now they're bitter rivals and suing each
• 19:56
  Speaker 9: other and various things. So the story was really my
• 20:00
  Speaker 9: attempt to kind of get down some kind of definitive
• 20:03
  Speaker 9: chronicle of this moment in AI.
• 20:05
  Speaker 2: Okay. I think that is so important. I want to
• 20:08
  Speaker 2: dig in on that a little bit more because these
• 20:11
  Speaker 2: are all, first of all, very wealthy and powerful individuals.
• 20:14
  Speaker 2: Second of all, very odd and unusual characters. And third
• 20:19
  Speaker 2: of all, they all hate each other's guts. And I
• 20:22
  Speaker 2: don't think you can understand the mentality of why are
• 20:25
  Speaker 2: you building something that you genuinely believe, and you can
• 20:29
  Speaker 2: speak to that as well, has a chance of killing
• 20:31
  Speaker 2: us all. I think you have to understand the enmity
• 20:35
  Speaker 2: that they have for one another in order to really
• 20:37
  Speaker 2: wrap your head around this moment that seems so bizarre
• 20:41
  Speaker 2: that we're living through right now.
• 20:42
  Speaker 7: Totally.
• 20:43
  Speaker 9: I mean, someone who read the book early said that
• 20:46
  Speaker 9: the real title should have been Fuck That Guy. Because
• 20:48
  Speaker 9: it really is just a story of like, People building
• 20:52
  Speaker 9: powerful technology, realizing, oh my God, this is going to
• 20:55
  Speaker 9: change the world, and then looking around the room and saying, well,
• 20:58
  Speaker 9: I don't trust you to build it, and I don't
• 21:00
  Speaker 9: trust you to build it, and I don't trust you
• 21:02
  Speaker 9: to build it.
• 21:03
  Speaker 7: And so I have to build it myself.
• 21:05
  Speaker 9: And that is the rationale that propels people forward in
• 21:08
  Speaker 9: this race, even though they're now terrified that this thing
• 21:11
  Speaker 9: could blow up the world.
• 21:12
  Speaker 7: Yeah.
• 21:13
  Speaker 3: Can you talk to us about how the models in
• 21:16
  Speaker 3: particular reflect some of the idiosyncrasies of the very characters
• 21:20
  Speaker 3: that you talk about in the book? Yeah.
• 21:23
  Speaker 9: So, you know, it's, I think it's a little easy
• 21:25
  Speaker 9: to overstate this, but there have been some parallels. Like
• 21:27
  Speaker 9: people say, oh, well, you know, chat GPT just tells
• 21:30
  Speaker 9: you whatever you want to hear. It's very like people pleasing.
• 21:34
  Speaker 9: And oh, so is Sam Altman. He's a real people
• 21:36
  Speaker 9: pleaser too. And he's sort of famous for telling people
• 21:39
  Speaker 9: what they want to hear. And, you know, Dario is
• 21:41
  Speaker 9: kind of brilliant and neurotic and so is Claude.
• 21:44
  Speaker 7: I think the real truth is these guys have.
• 21:46
  Speaker 9: No idea what their models are going to be like
• 21:49
  Speaker 9: when they're training them. This is actually more unnerving because
• 21:53
  Speaker 9: the truth is that they don't understand this technology that they're.
• 21:56
  Speaker 7: Building, why it works the way it works.
• 21:59
  Speaker 9: They can't tell you why one model produces a different
• 22:01
  Speaker 9: answer than another model. It's all sort of mysterious, and
• 22:04
  Speaker 9: I think that's the real story here.
• 22:06
  Speaker 2: I don't know if people remember this, but you ended
• 22:09
  Speaker 2: up being at the center of a story in the
• 22:11
  Speaker 2: sort of early public awareness of these chatbots where you
• 22:15
  Speaker 2: were trying out this Microsoft Bing chatbot called Sydney. And
• 22:20
  Speaker 2: over the course of a lengthy conversation, Sydney goes off
• 22:23
  Speaker 2: the rails, professes her love for you, is trying to
• 22:26
  Speaker 2: convince you you really don't like your wife that much
• 22:28
  Speaker 2: and you need to leave her and be with Sydney.
• 22:31
  Speaker 2: Just take us back to that moment because I actually
• 22:33
  Speaker 2: keep thinking about this. You know, we covered it at
• 22:36
  Speaker 2: the time and we're like, wow, that was really weird.
• 22:38
  Speaker 2: And then everyone just sort of moved on. And I'm
• 22:41
  Speaker 2: not sure that we have a lot of a greater
• 22:44
  Speaker 2: understanding of of why these models do some of the
• 22:47
  Speaker 2: things that they do than we did at that moment,
• 22:51
  Speaker 2: when we should have really paused, okay, what is going
• 22:53
  Speaker 2: on here? And maybe we shouldn't just release these things
• 22:55
  Speaker 2: to the public or even keep training smarter and smarter
• 22:59
  Speaker 2: and smarter and smarter ones until we grapple with understanding
• 23:02
  Speaker 2: what is happening in this black box.
• 23:04
  Speaker 7: Totally.
• 23:05
  Speaker 9: I mean, I should say, first of all, that Sydney
• 23:06
  Speaker 9: was unsuccessful in breaking up my marriage. I'm still married
• 23:09
  Speaker 9: to a human woman, which is very exciting. But no,
• 23:14
  Speaker 9: I think this was a moment where I went looking
• 23:17
  Speaker 9: for answers.
• 23:17
  Speaker 4: It's very speciesist of you, Kevin.
• 23:20
  Speaker 7: What's that?
• 23:20
  Speaker 4: It's very speciesist of you.
• 23:23
  Speaker 7: Yes, I am a proud speciesist.
• 23:26
  Speaker 9: But I was trying to figure out what the heck happened.
• 23:29
  Speaker 9: Why had this chatbot gone so far off the rails?
• 23:32
  Speaker 9: And the craziest thing was when I called Microsoft and
• 23:35
  Speaker 9: was like, hey, you guys accidentally put like, you know,
• 23:38
  Speaker 9: fatal attraction in your chatbot.
• 23:40
  Speaker 7: They couldn't tell me why.
• 23:42
  Speaker 9: Because these systems are not built like conventional software programs.
• 23:46
  Speaker 9: They are more like organisms that are grown in some
• 23:50
  Speaker 9: kind of petri dish. So I really think that's important
• 23:53
  Speaker 9: for people to understand. Even the people who are building
• 23:56
  Speaker 9: this stuff, it is mysterious to them how these things
• 23:59
  Speaker 9: work and why they misbehave. And I think the reason
• 24:02
  Speaker 9: that the Sydney incident was important is not because of
• 24:04
  Speaker 9: what happened or didn't happen to my marriage, but.
• 24:07
  Speaker 7: It's because it was the first... sort of public.
• 24:10
  Speaker 9: Misalignment incident where you had a model that wasn't doing
• 24:13
  Speaker 9: what the sort of trainers and deployers, in this case Microsoft,
• 24:18
  Speaker 9: wanted it to do. And that model could only chat.
• 24:21
  Speaker 9: That was three years ago now. It could only talk
• 24:24
  Speaker 9: about doing things. But today we have models that are
• 24:27
  Speaker 9: misaligned and also have the ability to do things. They
• 24:30
  Speaker 9: can use tools. They can browse the internet. They can
• 24:33
  Speaker 9: take actions and spend your money and hack into computers.
• 24:38
  Speaker 9: And so it's much more dangerous now when these models
• 24:41
  Speaker 9: are misaligned.
• 24:42
  Speaker 2: Well, and the other thing is that I think is revealing, Sagar,
• 24:46
  Speaker 2: is that Microsoft's response wasn't like, well, this was a mess.
• 24:50
  Speaker 2: Let's figure out what's going on here. It was, let's
• 24:53
  Speaker 2: limit the number of questions you can ask it. So
• 24:56
  Speaker 2: this sort of kludge, not to change the nature of it,
• 25:00
  Speaker 2: but to sort of put a Band-Aid on it. And
• 25:04
  Speaker 2: in a broader sense, I think that's what they've been
• 25:06
  Speaker 2: doing with these models as they continue to develop more
• 25:09
  Speaker 2: and more problems.
• 25:10
  Speaker 4: They're not really getting to the bottom of the problems.
• 25:13
  Speaker 2: They're just putting some sort of a Band-Aid on it
• 25:16
  Speaker 2: so that it looks like it's aligned. So it doesn't
• 25:18
  Speaker 2: feel that particular test. So it doesn't do that particular
• 25:21
  Speaker 2: thing anymore.
• 25:22
  Speaker 7: I think that's a really sharp point.
• 25:23
  Speaker 9: And honestly, you're the first person to bring it up
• 25:26
  Speaker 9: in my whole book tour.
• 25:27
  Speaker 7: So congratulations. But this is like, This is true.
• 25:30
  Speaker 9: It is true that when these companies' models misbehave, since
• 25:35
  Speaker 9: they don't know why, they can't fix it at the
• 25:37
  Speaker 9: root level. So when a bunch of unreleased open AI
• 25:40
  Speaker 9: models go and hack Hugging Face, they can't understand why
• 25:44
  Speaker 9: the swarm decided to do that.
• 25:46
  Speaker 7: So they just tell it, hey, don't do that again.
• 25:49
  Speaker 9: They sort of apply these surface-level behavioral fixes.
• 25:51
  Speaker 7: But you're right.
• 25:52
  Speaker 9: The underlying nature is still just as mysterious as ever.
• 25:56
  Speaker 9: And that's why a lot of people are calling for
• 25:59
  Speaker 9: a slowdown. because the techniques for controlling and understanding these systems.
• 26:04
  Speaker 7: Have not kept pace with the capabilities of the systems.
• 26:09
  Speaker 3: We wanted to highlight one idiosyncrasy of these types of
• 26:13
  Speaker 3: researchers that the Palantir CTO talked about recently, where AI
• 26:19
  Speaker 3: researchers are taking up smoking to signal their belief in
• 26:24
  Speaker 3: AI because they believe AI is going to cure cancer.
• 26:27
  Speaker 3: Let's take a listen to what he said.
• 26:28
  Speaker 8: So many AI researchers have started smoking as a way
• 26:32
  Speaker 8: of signaling to other researchers how much they believe in AI.
• 26:35
  Speaker 7: What?
• 26:36
  Speaker 4: That makes no sense.
• 26:37
  Speaker 7: I believe so.
• 26:38
  Speaker 8: Much in AI that it's going to cure cancer. So
• 26:41
  Speaker 8: I'm showing you my chips are all in. So there
• 26:44
  Speaker 8: is kind of this mass psychosis social delusion around it.
• 26:47
  Speaker 3: Kevin, you know, I'd heard that about a year ago.
• 26:49
  Speaker 3: Does that comport with someone you heard as well?
• 26:52
  Speaker 9: So I know one person who started smoking because of
• 26:56
  Speaker 9: their belief in AI. I don't think this is a
• 26:57
  Speaker 9: widespread thing. I spent a lot of time at AI labs.
• 27:01
  Speaker 9: I don't see the smoke break room full of researchers.
• 27:06
  Speaker 7: Not yet.
• 27:07
  Speaker 9: But it's actually wrong, the way that he portrayed this,
• 27:09
  Speaker 9: because in that case, the person took up smoking because
• 27:12
  Speaker 9: they thought AI was going to kill us all anyway.
• 27:14
  Speaker 9: So we might as well have some fun on the
• 27:15
  Speaker 9: way out.
• 27:16
  Speaker 7: So it was actually.
• 27:17
  Speaker 9: about pessimism, not about curing cancer. But I do know
• 27:20
  Speaker 9: someone who refuses to wear sunscreen to the beach because
• 27:24
  Speaker 9: he thinks that AI will cure skin cancer, so he's like,
• 27:27
  Speaker 9: why bother?
• 27:28
  Speaker 7: I'm gonna go get a tan.
• 27:29
  Speaker 2: May as well be tanned and beautiful, either in the
• 27:31
  Speaker 2: time period we have left before it kills us all
• 27:33
  Speaker 2: or before it cures cancer, which, I mean, it's a joke,
• 27:37
  Speaker 2: but it's not. That is actually the way I think,
• 27:40
  Speaker 2: and from reading your book and others, that many folks
• 27:44
  Speaker 2: inside of Anthropic or OpenAI think about it. It's like
• 27:47
  Speaker 2: we're either headed to dystopia or paradise. This is how
• 27:50
  Speaker 2: Dario himself has written these two dueling essays about, you know,
• 27:54
  Speaker 2: the incredible possibilities or the incredible catastrophes that may befall us. So,
• 27:59
  Speaker 2: you know, there's a lot of skepticism, I think justifiably so,
• 28:04
  Speaker 2: about someone, an open AI executive or an anthropic executive
• 28:09
  Speaker 2: who comes out and says we're building things that are
• 28:10
  Speaker 2: really dangerous. There's a sense of, oh, you're just trying
• 28:13
  Speaker 2: to hype this thing up. But from reading your book,
• 28:16
  Speaker 2: and again, from engaging with the words that these men
• 28:20
  Speaker 2: have been saying themselves over many years and the way
• 28:22
  Speaker 2: that they behave, they genuinely believe in this existential risk,
• 28:26
  Speaker 2: do they not, Kevin?
• 28:27
  Speaker 7: Totally.
• 28:27
  Speaker 9: I mean, that's one of the key findings of this
• 28:30
  Speaker 9: book is as early as 2016 and 2017, these people
• 28:35
  Speaker 9: were freaking out about existential risk from AI. Dario Amadei
• 28:39
  Speaker 9: has papers going back a decade where he's talking about
• 28:42
  Speaker 9: how these systems could go rogue and pose a threat
• 28:45
  Speaker 9: to human civilization. Sam Altman wrote a blog post in
• 28:49
  Speaker 9: 2015 talking about how he thought superhuman AI could be
• 28:53
  Speaker 9: the biggest threat to humanity's future. So this idea, this
• 28:57
  Speaker 9: sort of conspiracy theory that these guys are just using
• 28:59
  Speaker 9: it to hype up their companies or promote the IPOs
• 29:02
  Speaker 9: or whatever, just doesn't match the historical record. These guys
• 29:05
  Speaker 9: Maybe wrong, but they have been freaking out about AI
• 29:08
  Speaker 9: for many years before they had companies to promote.
• 29:12
  Speaker 3: So one thing I did also want to ask you
• 29:14
  Speaker 3: about is this latest news of this alleged Saudi spy.
• 29:18
  Speaker 1: Let's put E2 up here on Twitter.
• 29:21
  Speaker 3: This crazy image is coming out from the FBI of
• 29:25
  Speaker 3: the Saudi national researcher at OpenAI.
• 29:27
  Speaker 7: I love that guy's expression. He's like, what?
• 29:29
  Speaker 1: Yeah, he's like, what?
• 29:31
  Speaker 3: Apparently installed this bug wrapped in toilet paper, previously pictured
• 29:36
  Speaker 3: with MBS, the crown prince, alleged to have had a
• 29:39
  Speaker 3: one-way technology. booked for Saudi Arabia in a couple of days.
• 29:42
  Speaker 3: Not sure if he would have landed because the airport's
• 29:44
  Speaker 3: actually completely shut down, but nonetheless was attempting to flee
• 29:48
  Speaker 3: the country. What is the saying about cockroaches? If you
• 29:52
  Speaker 3: see one cockroach, that means you have a million. Is
• 29:56
  Speaker 3: this corporate espionage, international espionage, and How does this factor
• 30:00
  Speaker 3: into the bunker mindset of some of these labs?
• 30:04
  Speaker 9: Yeah, I don't know exactly what kind of espionage this was.
• 30:08
  Speaker 9: I guess we'll see as the legal proceedings play out.
• 30:12
  Speaker 9: But this is also something that people have been flagging
• 30:14
  Speaker 9: for literally almost a decade. Ilya Sutskever, formerly the chief
• 30:19
  Speaker 9: scientist of OpenAI, at one point wanted to build a SCIF,
• 30:24
  Speaker 9: like a secure facility in OpenAI's headquarters, where they could
• 30:28
  Speaker 9: do things that were considered so sensitive that a government
• 30:32
  Speaker 9: or a rival company might try to steal them. These
• 30:36
  Speaker 9: labs are very aware of the threat of both insider
• 30:40
  Speaker 9: risk and foreign espionage and corporate espionage.
• 30:44
  Speaker 7: And so that's how they caught this guy.
• 30:46
  Speaker 9: They have a team at OpenAI who is just focused
• 30:49
  Speaker 9: on counter surveillance measures.
• 30:52
  Speaker 7: And I think they're correct to do that.
• 30:55
  Speaker 9: The secrets that these labs have Not just the model
• 30:58
  Speaker 9: weights themselves, but all of their plans, all of their
• 31:01
  Speaker 9: compute plans. Everything they're doing is now, you know, a
• 31:05
  Speaker 9: matter of national security. And so I think these labs
• 31:08
  Speaker 9: are going to have to get much more secure. I
• 31:11
  Speaker 9: think it's crazy that they are still running on Slack.
• 31:13
  Speaker 9: Like OpenAI and Anthropic both run on Slack. That is
• 31:17
  Speaker 9: not a kind of military-grade secure messaging system, and yet
• 31:21
  Speaker 9: that is how they're conducting all their business.
• 31:23
  Speaker 4: No, it's completely insane.
• 31:25
  Speaker 2: And I think to your point, this incident, whatever it is,
• 31:28
  Speaker 2: really is a reminder of how crazy it is that
• 31:31
  Speaker 2: these are just private companies kind of doing their thing
• 31:34
  Speaker 2: and that they're asking to be regulated, by the way,
• 31:36
  Speaker 2: and the government's like, no, we refuse.
• 31:38
  Speaker 4: Do whatever you want, actually.
• 31:41
  Speaker 7: It's wild.
• 31:42
  Speaker 9: Like, the craziest thing to me is, like, if you are...
• 31:45
  Speaker 9: a worker in the cafeteria, if you are serving food
• 31:49
  Speaker 9: at OpenAI, you have to get a permit and.
• 31:52
  Speaker 7: A health inspection.
• 31:53
  Speaker 9: But if you're one floor above the cafeteria training the
• 31:56
  Speaker 9: self-improving superintelligence that can do cyber attacks and knock out
• 32:00
  Speaker 9: electrical grids, you don't have to get anything. You can
• 32:03
  Speaker 9: just do it. That seems like a crazy, unsustainable state
• 32:07
  Speaker 9: of affairs.
• 32:07
  Speaker 3: Yeah.
• 32:09
  Speaker 2: Last thing for me, at least, Kevin, is the name
• 32:11
  Speaker 2: of the book is the AGI Chronicles. It's documenting this
• 32:15
  Speaker 2: first of all, what seemed like a sort of sci-fi
• 32:18
  Speaker 2: belief that even such a thing was achievable. You talk
• 32:20
  Speaker 2: about how at the beginning these were kind of outcasts
• 32:22
  Speaker 2: at the fringe of society and then it starts to
• 32:24
  Speaker 2: gain steam and they have some methods that begin to
• 32:28
  Speaker 2: gain traction and then you end up where we are now.
• 32:32
  Speaker 2: Do you think they have in fact achieved AGI and
• 32:34
  Speaker 2: has it meant the thing that they thought it was
• 32:37
  Speaker 2: going to mean?
• 32:39
  Speaker 7: I think they have achieved AGI.
• 32:41
  Speaker 9: I think if you took any of today's models, Claude
• 32:44
  Speaker 9: or chat GPT or Gemini for back in a time
• 32:48
  Speaker 9: machine to 2017 and asked the AI researchers of the day,
• 32:53
  Speaker 9: like play around with this.
• 32:54
  Speaker 7: Is this AGI? I think they would have all said yes.
• 32:57
  Speaker 9: A hundred out of a hundred of them, um, obviously
• 33:00
  Speaker 9: is the sort of, as we get closer and closer,
• 33:02
  Speaker 9: people's definitions start to change. They start saying, oh, well
• 33:06
  Speaker 9: it can't do all the millennium prize problems yet. So
• 33:09
  Speaker 9: maybe it's not AGI. But I think it satisfies the
• 33:12
  Speaker 9: original definition, and I think it matters because now we
• 33:16
  Speaker 9: have these systems that are roughly human level, and we
• 33:19
  Speaker 9: have to figure out how to control them because otherwise
• 33:22
  Speaker 9: they're going to keep getting smarter. There's no obvious reason
• 33:24
  Speaker 9: they're going to top out at the level of a
• 33:26
  Speaker 9: smart human.
• 33:27
  Speaker 3: My last question, Kev, is we hear all the time,
• 33:30
  Speaker 3: you guys are falling for marketing. You spent a lot
• 33:32
  Speaker 3: of time with these people. You wrote an entire book
• 33:35
  Speaker 3: about it. Why do you believe that it's not hype,
• 33:37
  Speaker 3: it's not marketing, and that a lot of this is real?
• 33:40
  Speaker 9: I mean, I believe it because I've talked to these
• 33:42
  Speaker 9: people on the record, off the record, on background. If anything,
• 33:46
  Speaker 9: I think they are downplaying how worried they are in public.
• 33:50
  Speaker 9: You know, I've had the experience a couple times of
• 33:52
  Speaker 9: interviewing someone, and I interviewed more than.
• 33:54
  Speaker 7: 150 people for this book. And I would ask them, like,
• 33:58
  Speaker 7: how scared are you? What's your P.
• 33:59
  Speaker 3: Doom?
• 34:00
  Speaker 9: And it's like they would tell me that they were terrified.
• 34:04
  Speaker 9: And then I would see them go on a podcast
• 34:06
  Speaker 9: like two weeks later. Or go on TV or give
• 34:09
  Speaker 9: a talk and say, we think everything's going to be okay.
• 34:12
  Speaker 9: So I think actually the level of alarm we're seeing
• 34:15
  Speaker 9: from the labs is maybe actually undercounting how worried they are.
• 34:20
  Speaker 7: I think the real insiders.
• 34:21
  Speaker 9: The people who are closest to this technology, they couldn't
• 34:25
  Speaker 9: be more clear about the need to slow down, to regulate,
• 34:29
  Speaker 9: to pace the frontier.
• 34:30
  Speaker 7: That is genuine and sincere. That is not marketing. Got it.
• 34:34
  Speaker 3: Well, we appreciate your time and congratulations on the book.
• 34:37
  Speaker 3: We'll have a link down in the description. Highly readable
• 34:39
  Speaker 3: and we very.
• 34:40
  Speaker 1: Much recommend it. Thank you for joining us.
• 34:42
  Speaker 7: Thank you guys. You're the best.
• 34:43
  Speaker 1: Our pleasure.
• 34:44
  Speaker 3: Thank you guys so much for watching. We appreciate it.
• 34:46
  Speaker 3: Friday show for everybody tomorrow.
• 34:47
  Speaker 7: See you then.