Latest / Breaking Points with Krystal and Saagar / 6/25/26: Trump Holds Housing Bill Hostage, Trump Freaks Over War Powers, NSPM-7 Crackdown, AI Hype Debate

Transcript

6/25/26: Trump Holds Housing Bill Hostage, Trump Freaks Over War Powers, NSPM-7 Crackdown, AI Hype Debate · Breaking Points with Krystal and Saagar

• 0:00
  Speaker 1: Hey, guys, Saga and Crystal here.
• 0:01
  Speaker 2: Independent media just played a truly massive role in this election,
• 0:05
  Speaker 2: and we are so excited about what that means for
• 0:07
  Speaker 2: the future of this show.
• 0:08
  Speaker 1: This is the only place where you can find honest
• 0:10
  Speaker 1: perspectives from the left and the right that simply does
• 0:13
  Speaker 1: not exist anywhere else.
• 0:14
  Speaker 2: So if that is something that's important to you, please
• 0:17
  Speaker 2: go to Breakingpoints dot com. Become a member today and
• 0:19
  Speaker 2: you get access to our full shows, unedited, ad free,
• 0:22
  Speaker 2: and all put together for you every morning in your inbox.
• 0:25
  Speaker 1: We need your help to build the future of independent
• 0:27
  Speaker 1: news media and we hope to see you at Breakingpoints
• 0:29
  Speaker 1: dot com.
• 0:30
  Speaker 2: So Trump had a bit of a meltdown yesterday on
• 0:33
  Speaker 2: True Social and announced that he is going to refuse
• 0:37
  Speaker 2: to sign some legislation that pass in bipartisan manner that
• 0:41
  Speaker 2: is meant to promote housing affordability, so to bring down
• 0:44
  Speaker 2: housing costs in less and until they pass his Save
• 0:48
  Speaker 2: America Act, which is about sort of like vote rigging
• 0:51
  Speaker 2: in Republicans favor. We can get into war the details
• 0:53
  Speaker 2: in that. In just a moment, he was asked by
• 0:55
  Speaker 2: the press, why would you do this, why would you
• 0:57
  Speaker 2: hold housing affordability hostage in favor of this will charitably
• 1:02
  Speaker 2: call it voter reform legislation that they're putting. Let's take
• 1:04
  Speaker 2: a listen to. His response by the.
• 1:06
  Speaker 3: Moment is about a table.
• 1:07
  Speaker 2: We're so many Americans and this helped excellent say more
• 1:10
  Speaker 2: for a to you.
• 1:11
  Speaker 4: And flaying House.
• 1:13
  Speaker 3: It's just every election is important. We're doing very well.
• 1:17
  Speaker 3: They want a lot of communists to come in.
• 1:19
  Speaker 2: I'm saying it a little bit differently, but the people
• 1:22
  Speaker 2: that they're pushing are communists. So he just starts talking
• 1:26
  Speaker 2: about communist winning and taking over, etc. I guess that's
• 1:29
  Speaker 2: his rationale for why this legislation is so important to him.
• 1:32
  Speaker 2: Let's put see two up on the screen. This is
• 1:34
  Speaker 2: one of the true socials he put out announcing this decision.
• 1:37
  Speaker 2: Today's housing news conference and signing is hereby canceled until
• 1:41
  Speaker 2: such time as we passed the desperately needed Save America Act,
• 1:43
  Speaker 2: which I consider to be a national emergency. He put
• 1:46
  Speaker 2: out another one. Let's put this up on the screen,
• 1:49
  Speaker 2: the Elizabeth Pocahonas Warren centric housing bill, which is of
• 1:54
  Speaker 2: minor importance compared to lower interest rates, and even Faiza
• 1:57
  Speaker 2: pales in comparison to passing this Save America Act. That
• 2:01
  Speaker 2: is what Americans, both Democrats, Republicans and everyone else care about.
• 2:06
  Speaker 2: I'm not sure the pulling reflects that one get the
• 2:08
  Speaker 2: bad Republicans to approve it, or, better yet, terminate the
• 2:11
  Speaker 2: filibuster and approve it and everything else Republicans have dreamed of.
• 2:16
  Speaker 2: The Democrats would do it in one hour, one hundred percent.
• 2:20
  Speaker 2: Republicans will feel very stupid if they don't do it first.
• 2:23
  Speaker 2: I'll be watching, he says Ryan with tears in my eyes.
• 2:30
  Speaker 3: What are we supposed to what are we even supposed
• 2:32
  Speaker 3: to do with this guy?
• 2:34
  Speaker 4: So yes, Elizabeth Warren is involved in the the cop
• 2:38
  Speaker 4: and the negotiations around this housing bill, but it is
• 2:41
  Speaker 4: a widely bipartisan thing.
• 2:44
  Speaker 2: With permissions key pit pushed in an executive order. That's
• 2:48
  Speaker 2: the way this happened. He's like, oh, I want to
• 2:50
  Speaker 2: do this thing a blocking permanent capital from buying houses.
• 2:53
  Speaker 2: Elizabeth Warren said, okay, bet let's do it. We'll make
• 2:56
  Speaker 2: it into legislation, everybody signs on it gets passed. Now.
• 2:58
  Speaker 2: It's like the Elizabeth would Pokehonta's bill is not important
• 3:02
  Speaker 2: Save America Act whatever, I mean, it's just insane and
• 3:05
  Speaker 2: to be like, the public is really clamoring for my
• 3:08
  Speaker 2: voter id bill and not housing affordability is like what
• 3:13
  Speaker 2: planet are you literally on?
• 3:15
  Speaker 4: He tried this with the Ken Ken Paxton, you know
• 3:19
  Speaker 4: he said, I'm I will well Ken Paxon kind of
• 3:22
  Speaker 4: use it again. Some say if you pass the Save
• 3:24
  Speaker 4: America Act, then I will drop out and endorse Cornyn.
• 3:29
  Speaker 4: They don't have the votes to pass this thing without
• 3:31
  Speaker 4: getting rid of the philibuster. They don't have the votes
• 3:34
  Speaker 4: to get rid of the filibuster. They just don't. And
• 3:37
  Speaker 4: like the idea that the Senate Republicans would cave on
• 3:43
  Speaker 4: that they want they want the filibuster.
• 3:45
  Speaker 3: There's a reason they're not getting rid of the filibuster.
• 3:47
  Speaker 3: They like it in place.
• 3:49
  Speaker 4: The idea that they're going to get rid of it
• 3:51
  Speaker 4: just to get him to sign that his own housing
• 3:54
  Speaker 4: bill is just completely crazy.
• 3:58
  Speaker 3: It's not going to happen.
• 3:59
  Speaker 4: So now he's going to have to I guess, I
• 4:03
  Speaker 4: guess he has until Taco Tuesday, Right, there's another there's
• 4:06
  Speaker 4: another one coming, so he can he can schedule the
• 4:08
  Speaker 4: signing on Tuesday, or he can go ahead and the
• 4:11
  Speaker 4: veto and Housing Affordability bill when it is housing unaffordability
• 4:21
  Speaker 4: that is creating this communist takeover that he's complaining about.
• 4:24
  Speaker 3: It's not a communist takeover, but like.
• 4:26
  Speaker 4: The thing, the thing that is driving people in the
• 4:29
  Speaker 4: direction that he doesn't like, could be addressed, not completely,
• 4:33
  Speaker 4: but in significant part by this bill. And then he says, oh, well,
• 4:36
  Speaker 4: it's better to get interest rates down. It's like, buddy,
• 4:39
  Speaker 4: you got your guy onto the Federal Reserve. He is
• 4:44
  Speaker 4: a complete and total Trump flunky, but you have so
• 4:47
  Speaker 4: badly mismanaged the economy through your tariffs and your war
• 4:52
  Speaker 4: that even your guy is saying, yeah, I can't really
• 4:57
  Speaker 4: cut rates at this point because Trump is so the
• 5:00
  Speaker 4: economy that we're getting significant inflation that I have to
• 5:03
  Speaker 4: deal with. Meanwhile, he came in and said we're no
• 5:07
  Speaker 4: longer doing guidance about what we're going to do in
• 5:10
  Speaker 4: the future, and so Wall Street responded to that by saying, well, okay,
• 5:14
  Speaker 4: that means there's more risk. We don't know what's going
• 5:16
  Speaker 4: to happen, is more uncertainty. Uncertainty is priced in as
• 5:19
  Speaker 4: a risk premium, which increases interest rates. So interest rates
• 5:23
  Speaker 4: so he's his own guy with his new policy that
• 5:25
  Speaker 4: he put into place to the Fed is driving up
• 5:28
  Speaker 4: interest rates. So h Trump is just doing everything wrong
• 5:34
  Speaker 4: at every turn. It's absolutely remarkable.
• 5:37
  Speaker 2: Yeah, and just so people know, you know, the Save
• 5:39
  Speaker 2: America Act, we have covered it here before, but the
• 5:41
  Speaker 2: TLDR is, it's going to make it harder for people
• 5:44
  Speaker 2: to vote, and it requires states to send in their
• 5:47
  Speaker 2: voter roles. You know, a bunch of states. Georgia has
• 5:49
  Speaker 2: been sort of ground zero for this. They will automatically
• 5:54
  Speaker 2: kick a bunch of people off the rolls and force
• 5:56
  Speaker 2: you to prove that you are still a voter, that
• 5:58
  Speaker 2: you're still at this address. To get yourself readded. First,
• 6:01
  Speaker 2: you have to know that you were kicked off to
• 6:03
  Speaker 2: begin with. And they do all sorts of things like,
• 6:05
  Speaker 2: you know, if it's a duplicate name, or if it's
• 6:08
  Speaker 2: a name of someone who is undocumented who is not
• 6:12
  Speaker 2: a citizen, but this person just happens to have the
• 6:15
  Speaker 2: same name and is a citizen, they'll just auto delete
• 6:18
  Speaker 2: all those people off the rolls because they assume those
• 6:20
  Speaker 2: are more likely to be Democratic voters. And this is
• 6:22
  Speaker 2: sort of like a nationwide effort to do that. But
• 6:25
  Speaker 2: you know the other thing that is, let's put see
• 6:27
  Speaker 2: five up on the screen, because This is part and
• 6:29
  Speaker 2: parcel of his attempt to tilt the midterm playing field
• 6:34
  Speaker 2: strongly in the Republican direction. Most obvious effort in that
• 6:37
  Speaker 2: has been the whole redistricting push, which thanks to terrible
• 6:40
  Speaker 2: court decisions, has really skewed in favor of Republicans. Now
• 6:45
  Speaker 2: we have the Postmaster General saying that the US Postal
• 6:48
  Speaker 2: Service is not going to deliver mail ballots if states
• 6:52
  Speaker 2: don't give the Trump regime access to their voter rolls. Okay,
• 6:55
  Speaker 2: in many states right now are in litigation with the
• 6:59
  Speaker 2: Trump administry right now pushing back on this executive order
• 7:03
  Speaker 2: demand that they hand over their voter roles. So they're
• 7:06
  Speaker 2: trying to now weaponize the postal services say oh, well,
• 7:09
  Speaker 2: we're not going to deliver your ballots at all if
• 7:11
  Speaker 2: you don't do what we want you to do. And
• 7:13
  Speaker 2: so Ryan, there's been this question of whether Trump even
• 7:17
  Speaker 2: cares about the midterms because his political I guess actions
• 7:22
  Speaker 2: instincts have been so bad and so insane. You're like,
• 7:25
  Speaker 2: I guess he just doesn't care about politics at all anymore.
• 7:29
  Speaker 2: He's in his lame duck period already. He's just gonna
• 7:32
  Speaker 2: do whatever he feels like doing, and doesn't really give
• 7:34
  Speaker 2: a shit of Republicans lose the House. In the Senate,
• 7:37
  Speaker 2: these actions would tell a very different story that he
• 7:39
  Speaker 2: still is very interested in what can I do to
• 7:42
  Speaker 2: make sure that even if there is a public that
• 7:44
  Speaker 2: is dramatically opposed to what my administration does, that we're
• 7:48
  Speaker 2: still able to keep our grip on power.
• 7:50
  Speaker 4: Here, Yes, and some of the kind of resistance Democrats'
• 7:55
  Speaker 4: most hysterical claims about Trump during his first term, I
• 8:00
  Speaker 4: have really been born out over time. So I think
• 8:02
  Speaker 4: we have to really start, you know, we have to
• 8:05
  Speaker 4: take all of those claims very seriously. And the idea
• 8:10
  Speaker 4: that he wants to kind of break the link between popularity,
• 8:16
  Speaker 4: favorability ratings that the democratic will and what he can
• 8:19
  Speaker 4: do in office seems to be a rather credible one here,
• 8:23
  Speaker 4: So you think your point is right. It's not necessarily
• 8:25
  Speaker 4: that he doesn't care about the results of the elections.
• 8:27
  Speaker 4: It's more like he wants to kind of enact that
• 8:31
  Speaker 4: Stalin line where Joseph Stalin pamiously said it doesn't matter
• 8:34
  Speaker 4: who votes, it matters who counts the votes. In this case,
• 8:37
  Speaker 4: it would matter who delivers the ballots to the voters
• 8:41
  Speaker 4: and who has possession of.
• 8:43
  Speaker 3: The voter roles.
• 8:45
  Speaker 4: So I think Democrats are probably going to have to
• 8:50
  Speaker 4: take on a variety of his slogan back in twenty
• 8:54
  Speaker 4: twenty four, and they're going to have to win too
• 8:56
  Speaker 4: big to rig. It's going to have to be too overwhelming,
• 8:59
  Speaker 4: and perhaps he will lose so much political capital that
• 9:03
  Speaker 4: even the Postmaster General it's like, you know what, Actually,
• 9:06
  Speaker 4: I'm not going to kind of try to break American
• 9:09
  Speaker 4: democracy just to save this president who's at twenty eight
• 9:12
  Speaker 4: percent at this point. But then again, I think we
• 9:16
  Speaker 4: have to take everything.
• 9:17
  Speaker 2: Seriously, no doubt about it, no doubt about it. So
• 9:24
  Speaker 2: there's been another whole crazy situation unfolding with regard to
• 9:27
  Speaker 2: Trump and the Republicans. Actually, with regards to this Irun
• 9:30
  Speaker 2: War Powers resolution that actually passed the House past the Senate.
• 9:34
  Speaker 2: There were two Republicans, Mitch McConnell and McCormick, who were
• 9:38
  Speaker 2: both absent for it. Cormick was actually traveling with Trump
• 9:42
  Speaker 2: at the times it was vailable to go to the vote,
• 9:45
  Speaker 2: and then Ran Paul and Bill Cassidy voted in favor
• 9:49
  Speaker 2: of the war powers resolution for the first time this
• 9:52
  Speaker 2: thing passed. Trump was very very very very unhappy about this.
• 9:56
  Speaker 2: Reportedly can put d one up on the screen. He
• 9:59
  Speaker 2: and apparently had an all out shouting match. The way
• 10:03
  Speaker 2: CBS News describes it as a sitting down, no, you
• 10:06
  Speaker 2: sit down, shouting spat. During the closed door meeting with
• 10:11
  Speaker 2: the Senate Republicans, Trump was absolutely furious, you know, lighting
• 10:16
  Speaker 2: into Cassidy and Ran Paul and Lisa Murkowski as well
• 10:21
  Speaker 2: for their votes on this resolution. And you know, at
• 10:24
  Speaker 2: this point, you've got Ryan. The Iran war is kind
• 10:29
  Speaker 2: of on hold here as we do these negotiations, so
• 10:32
  Speaker 2: classic Congress move here to actually pass the resolution once
• 10:36
  Speaker 2: the war may be settled and over. But it's also
• 10:39
  Speaker 2: very comforting to me that Trump was so angry about
• 10:43
  Speaker 2: this going through. Is not a great sign of confidence
• 10:47
  Speaker 2: that he feels he's done with war, I guess at
• 10:49
  Speaker 2: this point.
• 10:51
  Speaker 3: Yeah, I don't have.
• 10:52
  Speaker 4: An explanation for why it is, but I mentioned this
• 10:54
  Speaker 4: a bunch before. Trump has always really cared about these
• 10:58
  Speaker 4: war powers resolution folks like he he credits them with
• 11:03
  Speaker 4: real power. Even during Trump won when there would be
• 11:07
  Speaker 4: war powers pushes in the House or Senate, particularly at
• 11:11
  Speaker 4: the time on the Yemen War and the US support
• 11:14
  Speaker 4: for the Saudi led Yemen war like, he would fight
• 11:18
  Speaker 4: hard against them, and he would change what he was
• 11:21
  Speaker 4: doing for after one of them passed, he stopped kind
• 11:24
  Speaker 4: of refueling Saudi fighter jets, which brought it brought to
• 11:28
  Speaker 4: a significant end the Saudi Arabia's ability to carry out
• 11:32
  Speaker 4: that war and helped and helped kind of end Saudi
• 11:35
  Speaker 4: Arabia's involvement. And so, even though he has no respect
• 11:39
  Speaker 4: for Congress, no respect for anything other than kind of
• 11:41
  Speaker 4: his day to day whims, for some reason, he does
• 11:44
  Speaker 4: respect war powers votes and so too. And so that
• 11:50
  Speaker 4: I think explains why he had this kind of blow
• 11:52
  Speaker 4: up shouting match. You know, somebody like Cassidy, who Trump
• 11:57
  Speaker 4: threw out of office, doesn't have a whole lot of
• 11:59
  Speaker 4: motivation to give him what he wants. But it's interesting
• 12:02
  Speaker 4: that after getting a briefing, I guess you know that
• 12:05
  Speaker 4: he was they're able to kind of move him on this.
• 12:09
  Speaker 4: He was one of the ones that flipped right.
• 12:11
  Speaker 2: Yeah, So just an update to my previous comments. It
• 12:14
  Speaker 2: was four Republicans that voted against it, Cassidy, Collins, Murkowski,
• 12:17
  Speaker 2: Ran Paul, then Mitch McConnell and Dave McCormick that missed
• 12:20
  Speaker 2: the vote that it passes by fifty to forty eight.
• 12:23
  Speaker 2: John Fetterman voted with the Republicans, because of course he did.
• 12:26
  Speaker 2: It doesn't even say which Democrat was the one Democrat.
• 12:28
  Speaker 2: But I'm just going to assume, oh, John Fvetterman, because
• 12:30
  Speaker 2: he's been voting against all of these things. And it
• 12:34
  Speaker 2: is significant because so you know, so people know how
• 12:36
  Speaker 2: this works. It doesn't require presidential signature. Once it passes
• 12:39
  Speaker 2: the House and it passes the Senate, then that's it.
• 12:43
  Speaker 2: You know, you are supposed to not go to war,
• 12:46
  Speaker 2: you're supposed to stop the war, and so it gives
• 12:48
  Speaker 2: you some power to go to the courts. Then if
• 12:50
  Speaker 2: Trump was to restart hostilities direct hostilities with a run
• 12:55
  Speaker 2: and say he can't do this, right, it gives them
• 12:58
  Speaker 2: some legal power now that Trump has blown through previous
• 13:02
  Speaker 2: legal limits. I have no doubt they would come up
• 13:04
  Speaker 2: with some theory of the case why it doesn't count
• 13:06
  Speaker 2: and we can do whatever we want more. Yeah, and
• 13:09
  Speaker 2: so Ryan, I kind of you know, I don't know
• 13:11
  Speaker 2: that he's that concerned about the legal implications here. I
• 13:13
  Speaker 2: think it's more of an ego thing that you had
• 13:15
  Speaker 2: this many Republicans vote against him which is why so
• 13:18
  Speaker 2: he freaks out yells at everybody who voted the wrong way. Cassidy,
• 13:23
  Speaker 2: one of the things that he had apparently said is
• 13:25
  Speaker 2: you haven't even briefed us on this war and you
• 13:27
  Speaker 2: expect us to go along with it. So some of
• 13:30
  Speaker 2: the members of the Trump administration then later in the
• 13:32
  Speaker 2: day reached out to Cassidy say all right, we'll find
• 13:34
  Speaker 2: come in for a briefing, brought him into the situation room,
• 13:36
  Speaker 2: gave him a whole briefing. He tweets on, oh, thank
• 13:39
  Speaker 2: you for the briefing, and then he changes his vote
• 13:41
  Speaker 2: and they do the same War Powers resolution vote again,
• 13:44
  Speaker 2: and this time he and Ran Paul flipped their votes
• 13:48
  Speaker 2: so that the thing and the two that weren't there
• 13:50
  Speaker 2: are there, and so it's then able to pass. Now
• 13:54
  Speaker 2: it's kind of meaningless because you already have the one
• 13:57
  Speaker 2: that you know, you already have the one that passed. Yeah,
• 14:00
  Speaker 2: sorry failed, You already have the one that passed. So
• 14:03
  Speaker 2: the fact that it failed this time doesn't like supersede
• 14:06
  Speaker 2: that or override it. It's just purely an ego play
• 14:10
  Speaker 2: for Trump to come and be like, no, I'm going
• 14:12
  Speaker 2: to bend you to my will. I'm going to force
• 14:14
  Speaker 2: you to see the world the way that I see
• 14:17
  Speaker 2: the world, and you are going to publicly demonstrate your
• 14:21
  Speaker 2: fealty to me and the fact that I was right
• 14:24
  Speaker 2: and you were wrong on this vote.
• 14:27
  Speaker 4: Yeah, it's it's yes, and I think you're right that,
• 14:32
  Speaker 4: Like if he relaunched the war, he would say, well,
• 14:35
  Speaker 4: it's a new war, and heg SATs even hinted at
• 14:37
  Speaker 4: that like, Okay, we're only allowed to do six days
• 14:39
  Speaker 4: of war.
• 14:39
  Speaker 3: Well we stopped and so now now we're doing new one.
• 14:45
  Speaker 4: But just materially, he doesn't have the capacity to do that,
• 14:48
  Speaker 4: it seems.
• 14:49
  Speaker 2: Like, Yeah, last thing on is Ryan, I'm curious for
• 14:51
  Speaker 2: your you and what your reporting suggests. So Cassidy lost
• 14:54
  Speaker 2: this primary, so he's a lame dog, so he has
• 14:56
  Speaker 2: no reason to like continue to be slavishly pro Trump.
• 15:00
  Speaker 2: Sort of embarrassing to me that he went into this
• 15:02
  Speaker 2: briefing and flipped his vote, Like how weak are you?
• 15:04
  Speaker 5: I mean?
• 15:05
  Speaker 2: And just to pause on him for a second, this guy,
• 15:08
  Speaker 2: I think his sin was he voted in favor of
• 15:11
  Speaker 2: the like January sixth impeachment. Since then, he's done everything
• 15:16
  Speaker 2: they've wanted him to do. He is doctor and so
• 15:19
  Speaker 2: he did not really want to vote for RFK Junior
• 15:21
  Speaker 2: to be the head of HHS, but he did it,
• 15:24
  Speaker 2: and RFK Junal, I'm totally pro vaccine and don't worry
• 15:27
  Speaker 2: about it, sir. And they had these private meetings and
• 15:29
  Speaker 2: there was all sorts of Susan Collins esque angst about
• 15:32
  Speaker 2: how concertned he is. But the end of the day,
• 15:35
  Speaker 2: he voted for him. He voted for all the Trump picks.
• 15:37
  Speaker 2: He's done everything Trump wanted to do, but it didn't matter.
• 15:40
  Speaker 2: Trump came after him in this primary and he gets,
• 15:42
  Speaker 2: you know, kicked down of office. And so then for
• 15:45
  Speaker 2: him to just still be bending the knee whenever he's
• 15:48
  Speaker 2: asked to is frankly humiliating. But you have others as
• 15:53
  Speaker 2: well who have lost their seats. John Cornyn is now
• 15:57
  Speaker 2: Actually he tweeted out at Wall Street Journal article I
• 15:59
  Speaker 2: don't know if you all this that was like, the
• 16:02
  Speaker 2: top goal scorer for the men's soccer team is a
• 16:05
  Speaker 2: citizen treasonship. Yeah, And so he tweeted out a Wall
• 16:09
  Speaker 2: Street Journal architle that was like, this is what makes
• 16:11
  Speaker 2: our soccer team in America great. He's been giving some
• 16:14
  Speaker 2: interviews that are more critical. You've got a few of these,
• 16:17
  Speaker 2: and Thomas Massey obviously in the House, so you do
• 16:19
  Speaker 2: have a few more of these characters who don't necessarily
• 16:23
  Speaker 2: have to go along to get along the way that
• 16:25
  Speaker 2: they used to. And you have Trump obviously with an
• 16:28
  Speaker 2: absolutely dismal approval rating and his political popularity falling. So
• 16:32
  Speaker 2: I don't know, it could make the next few months
• 16:34
  Speaker 2: slightly more interesting with slightly more descent than we've seen
• 16:37
  Speaker 2: in the past.
• 16:38
  Speaker 4: Yeah, and Trump still controls Donor Networks, he controls the IRS,
• 16:42
  Speaker 4: he controls the Justice Department like he's the most corrupt
• 16:45
  Speaker 4: president we've ever had and is willing to kind of
• 16:49
  Speaker 4: weaponize the institutional power that he has in the government.
• 16:53
  Speaker 4: So you know, if you're Cassidy, you're like, Okay, yeah,
• 16:57
  Speaker 4: I lost my race and I'm leaving office.
• 17:00
  Speaker 3: Guy still runs the DOJ, guy still runs the RS.
• 17:03
  Speaker 4: So maybe you know, if it's going to pass anyway
• 17:06
  Speaker 4: or it's going to fail anyway this next time around,
• 17:10
  Speaker 4: and he doesn't care that much, you know what, just
• 17:14
  Speaker 4: going to go along.
• 17:14
  Speaker 3: To get along. Actually, well, he knows what they said
• 17:16
  Speaker 3: in the situation room.
• 17:18
  Speaker 2: True, and the uh, the consequences of crossing Trump and
• 17:22
  Speaker 2: the grave the grave consequences that can result from dissenting
• 17:25
  Speaker 2: with this administration are a good segue actually to our
• 17:28
  Speaker 2: next block. So some extremely lengthy sentences were handed down
• 17:34
  Speaker 2: in a court in Texas for some anti ICE protesters.
• 17:38
  Speaker 2: Let's put the New York Times article up on the
• 17:40
  Speaker 2: screen that breaks down some of the details here. Protesters
• 17:43
  Speaker 2: accused of Antifa ties sentenced to up to one hundred
• 17:48
  Speaker 2: years in ICE attack. The subhad says the penalties issued
• 17:51
  Speaker 2: in an attack where a police officer was shot dwarf
• 17:54
  Speaker 2: those given to January six riders appeared to signal at
• 17:57
  Speaker 2: least some courts will deal aggressively with ICE protests. Let
• 18:00
  Speaker 2: me just read a little bit here so you can
• 18:02
  Speaker 2: get the details. The leader of a group of protesters
• 18:04
  Speaker 2: accused of being members of Antifa was sentenced on Tuesday
• 18:08
  Speaker 2: to one hundred years in prison after a jury found
• 18:11
  Speaker 2: him and seven other demonstrators guilty of supporting terrorism while
• 18:14
  Speaker 2: taking part in an armed assault last summer against an
• 18:17
  Speaker 2: immigration facility in Alvarado, Texas. The extraordinary sentence against the
• 18:21
  Speaker 2: protester Benjamin Song, was only one of the harsh penalties
• 18:24
  Speaker 2: meted out in the case. During separate hearings in Federal
• 18:27
  Speaker 2: District Court in Fort Worth by two judges who castigated
• 18:30
  Speaker 2: the defendants for using violence and attacking the democratic process
• 18:33
  Speaker 2: during the protest. Nine young demonstrators, including mister Song, were
• 18:37
  Speaker 2: found guilty in March of an array of charges seending
• 18:40
  Speaker 2: from the attack on the ice facility. Six of the
• 18:42
  Speaker 2: defendants were convicted of terrorism charges, along with mister Song,
• 18:46
  Speaker 2: sentenced to between fifty and seventy years. Another was found
• 18:48
  Speaker 2: guilty of lesser crimes, was not even present at the protest,
• 18:53
  Speaker 2: and was given a term of thirty years in prison.
• 18:56
  Speaker 2: A final defendant is scheduled to be sentenced next month.
• 19:00
  Speaker 2: Joining us now to break down this case and in
• 19:02
  Speaker 2: particular these incredibly lengthy sentences, how the federal government made
• 19:07
  Speaker 2: these charges stick and what they argued in order to
• 19:10
  Speaker 2: secure these thirty, fifty and one hundred year sentences is Pisco.
• 19:14
  Speaker 2: He is both YouTuber at the channel Piasco littie and
• 19:17
  Speaker 2: also a lawyer. So great to see you at Pisco.
• 19:20
  Speaker 6: Thank you for having me.
• 19:21
  Speaker 2: Yeah, So start with what was this protest? It seems
• 19:25
  Speaker 2: like the government argued that they all conspired as part
• 19:28
  Speaker 2: of an Antifa cell to go and try to murder
• 19:30
  Speaker 2: a law enforcement officer, is that, in fact what happened
• 19:33
  Speaker 2: to you're based on the evidence.
• 19:34
  Speaker 6: So there's a lot of disputes actually about sort of
• 19:37
  Speaker 6: their participation in Antifa. There's disputes even among the government's
• 19:41
  Speaker 6: own witnesses about whether they in fact were members of Antifa,
• 19:45
  Speaker 6: But certainly a lot of sort of left wing sympathies,
• 19:47
  Speaker 6: and we're talking about the affiliations of the group. But
• 19:49
  Speaker 6: in some and substance they're accused of on July fourth,
• 19:53
  Speaker 6: at around ten o'clock that night, going to the Prayerland
• 19:57
  Speaker 6: detention Center in alvar Alvarado Tech This to shoot off fireworks,
• 20:03
  Speaker 6: And that's where the kind of divergence occurs on the facts.
• 20:06
  Speaker 6: According to the protesters, that was all that they were
• 20:09
  Speaker 6: there to do was to shoot off fireworks and to
• 20:11
  Speaker 6: have like a noise demonstration to inspire the members of
• 20:15
  Speaker 6: the you know, who are being incarcerated there in the
• 20:17
  Speaker 6: detention center. But the government alleges that it was a
• 20:20
  Speaker 6: lot more of like a corrupt motive and a motive
• 20:22
  Speaker 6: to in fact, like inside a riot and potentially even
• 20:25
  Speaker 6: to ambush members of law enforcement.
• 20:27
  Speaker 5: But we can get into the weeds.
• 20:30
  Speaker 3: Well, yeah, let's get it let's get.
• 20:32
  Speaker 4: Into the weeds, like what evidence do they have, I mean,
• 20:36
  Speaker 4: they brought up getting convictions here, what evidence do they
• 20:39
  Speaker 4: have that they were trying to organize some sort of
• 20:43
  Speaker 4: like ambush or right, Like, having covered a lot of
• 20:46
  Speaker 4: left wing groups and protests in the past, the idea
• 20:50
  Speaker 4: that they would show up and set off some fireworks
• 20:53
  Speaker 4: and that would be the goal of the protest sounds
• 20:58
  Speaker 4: entirely plausible and accurate like that. That is what you know,
• 21:02
  Speaker 4: groups of an activists like this do like it would
• 21:04
  Speaker 4: not be at all strained. They like, let's do some fireworks,
• 21:07
  Speaker 4: brings some atension here, it's July fourth, is America, Uh,
• 21:13
  Speaker 4: civil disobedience, it'll be it's a fine. But the idea
• 21:16
  Speaker 4: that they're going to set up an ambush and start
• 21:18
  Speaker 4: killing people, like what so? What in the weeds, Like
• 21:21
  Speaker 4: what evidence do they have or do they or did
• 21:24
  Speaker 4: they present that that there was some other you know,
• 21:27
  Speaker 4: murderous motive here.
• 21:29
  Speaker 6: Yeah, So first is the fact that some of them
• 21:32
  Speaker 6: brought guns to the to the actual side of the protests,
• 21:35
  Speaker 6: and there was in Texas the ringleader Yeah in Texas, right, Well,
• 21:39
  Speaker 6: and you get into a lot of these conversations too,
• 21:41
  Speaker 6: about alex pretty and whether or not it's you know,
• 21:45
  Speaker 6: de facto unreasonable or de facto criminal intent to show
• 21:47
  Speaker 6: up to a protest, you know, with with guns and
• 21:50
  Speaker 6: so open caring, or or do anything like that. So
• 21:53
  Speaker 6: the evidence that the government relied on is number one,
• 21:56
  Speaker 6: that what they're wearing, right, they make a lot of
• 21:58
  Speaker 6: the fact that they're in black block, that the fact
• 22:00
  Speaker 6: that they all have code names, that they turned off
• 22:02
  Speaker 6: their cell phones in advance of what was happening here,
• 22:04
  Speaker 6: that they brought guns in the first place, that they
• 22:06
  Speaker 6: had like medical kits there, presumably to treat any injuries
• 22:10
  Speaker 6: that could occur. So on the one hand, you have like, well,
• 22:11
  Speaker 6: why are you bringing guns here? One of them actually
• 22:14
  Speaker 6: did fire towards an officer, and he was the one
• 22:17
  Speaker 6: who was convicted with attempted murder. But on the other hand,
• 22:20
  Speaker 6: you have no sort of messages from these various signal
• 22:23
  Speaker 6: groups that seem to indicate at least explicitly that their
• 22:26
  Speaker 6: intent was to ambush members of ice ore of the
• 22:30
  Speaker 6: police department who later on came. So there's a what
• 22:33
  Speaker 6: the defense might point to is like a lack of
• 22:35
  Speaker 6: corroborating evidence of explicitly violent motive within the messages that occurred.
• 22:40
  Speaker 6: There's one of the attorney also points to is the
• 22:43
  Speaker 6: fact that there's not even like messages about going and
• 22:45
  Speaker 6: vandalizing property, which they also did, and so you look
• 22:49
  Speaker 6: at the absence there, but then also the fact that
• 22:52
  Speaker 6: they you know, if their goal was to ambush, why
• 22:55
  Speaker 6: weren't they more well armed? Why wasn't this better coordinated?
• 22:58
  Speaker 6: And there start to be a little bit of a
• 23:00
  Speaker 6: repeat of some of the arguments levied regarding January sixth
• 23:04
  Speaker 6: about whether or not it's plausible that their intent here
• 23:06
  Speaker 6: was to get them to come out just to kill them.
• 23:09
  Speaker 6: If that were the case, then you know, why wasn't
• 23:11
  Speaker 6: it quote unquote more successful.
• 23:14
  Speaker 2: Okay, so you've got in one category gory, you have
• 23:17
  Speaker 2: the protester who did fire weapon at an officer, the
• 23:20
  Speaker 2: officers hit. That's the one that's getting one hundred years.
• 23:23
  Speaker 2: At the other end of the spectrum, and what I
• 23:25
  Speaker 2: think will lead any reasonable person to look at this
• 23:28
  Speaker 2: and go, this is insane. You have someone convicted and
• 23:32
  Speaker 2: sentenced to thirty years in prison here who was not
• 23:35
  Speaker 2: even at the protest. Let's put this Guardian article up
• 23:38
  Speaker 2: on the screen, so they say the ninth defendant, Daniel
• 23:41
  Speaker 2: Sanchez Estrata, was not at the protest, but was convicted
• 23:45
  Speaker 2: of corruptly concealing a document or record after prosecutors said
• 23:50
  Speaker 2: he moved left wing zines and other materials at the
• 23:54
  Speaker 2: request of his wife after she was arrested. He was
• 23:58
  Speaker 2: sentenced to thirty years in prison. So it says the
• 24:02
  Speaker 2: charge doesn't require prosecutors prove a connection to any kind
• 24:05
  Speaker 2: of terrorist ideology, only that a defendant provided support for
• 24:08
  Speaker 2: one of a list of several crimes. The Justice departments
• 24:11
  Speaker 2: spun the convictions as proof that Antifa was a terrorist organization.
• 24:15
  Speaker 2: So I want to repeat for everyone because I read
• 24:18
  Speaker 2: this like ten times to really process it and be like,
• 24:20
  Speaker 2: it can't be the case that he was convicted and
• 24:23
  Speaker 2: sentenced to thirty years in prison for moving some leftist literature.
• 24:27
  Speaker 2: But that is literally the case here. His wife was
• 24:31
  Speaker 2: at the protest, she gets arrested, she calls her husband,
• 24:34
  Speaker 2: he moves whatever this literature is, which, by the way,
• 24:37
  Speaker 2: they offered in evidence. You know some of these things
• 24:39
  Speaker 2: and it's, you know, kind of what you would expect.
• 24:41
  Speaker 2: There were things about immigrants' rights and how to stand
• 24:44
  Speaker 2: up to ice in these sorts of things. In any case,
• 24:47
  Speaker 2: you're allowed to you know, read whatever you want in
• 24:49
  Speaker 2: this country that's supposed to be the idea. And yet
• 24:51
  Speaker 2: for this action he was sentenced to thirty years in prison.
• 24:55
  Speaker 2: There were a number of the protesters who had already
• 24:58
  Speaker 2: when the security guards showed up and said you guys
• 25:00
  Speaker 2: got to go left before the weapon was ever even discharged.
• 25:05
  Speaker 2: There was the one protester who fired the weapon. There
• 25:09
  Speaker 2: were some who participate in acts of vandalism. There were
• 25:12
  Speaker 2: the fireworks, No one injured, just fireworks that were set off.
• 25:16
  Speaker 2: And now you have these sentences of thirty fifty and
• 25:18
  Speaker 2: one hundred years piece.
• 25:19
  Speaker 6: Go yeah, no, So, I mean, these are great points
• 25:22
  Speaker 6: that you're raising. It would be one thing if, okay,
• 25:25
  Speaker 6: a person is separate and apart from the protest activities
• 25:27
  Speaker 6: or let's say the planned ambush or whatever, and they're
• 25:29
  Speaker 6: like integral and planning the violent events. And you can
• 25:32
  Speaker 6: think about some of the conditionous or sorry seditious conspiracy
• 25:35
  Speaker 6: charges that were levied for example and Rico Tario and
• 25:38
  Speaker 6: other members there where they're integral and some of the
• 25:40
  Speaker 6: planning that's going on and where there's evidence of you
• 25:42
  Speaker 6: know that they're involved in the designs of the group
• 25:44
  Speaker 6: there on January sixth. But here this is defendant Sanchez,
• 25:48
  Speaker 6: I believe it. He wasn't even charged with providing material
• 25:50
  Speaker 6: support to terrorists. He was in charged with attempted murder.
• 25:54
  Speaker 6: He wasn't charged with setting off explosives, and so as
• 25:57
  Speaker 6: to him a thirty year prison sentence for this is
• 26:02
  Speaker 6: what amounts the obstruction of justice related offenses. For me,
• 26:05
  Speaker 6: it kind of is just totally disproportionate to what he's
• 26:09
  Speaker 6: alleged to be participating in, and it speaks to a
• 26:12
  Speaker 6: broader defficiency I think in the narrative that the government
• 26:15
  Speaker 6: is trying to push here, there were attempts to get
• 26:19
  Speaker 6: other members of this group, this alleged Antifa cell hit
• 26:22
  Speaker 6: with the attempted murder charges because of something called Pinkerton liability.
• 26:26
  Speaker 6: And this essentially says that if there is a conspiracy
• 26:29
  Speaker 6: and other members of the conspiracy are doing substantive crimes
• 26:33
  Speaker 6: that further the conspiracy, and they're like reasonably foreseeable and
• 26:35
  Speaker 6: lighted the conspiracy, you can hit them not just with
• 26:38
  Speaker 6: a conspiracy, but with like the substantive crimes that the
• 26:41
  Speaker 6: other members of the conspiracy did. It's called Pinkerton liability.
• 26:44
  Speaker 6: And here right, multiple members of this quote alleged Antifa
• 26:48
  Speaker 6: cell were charged with attempted murder, but the jury actually acquitted.
• 26:53
  Speaker 6: They acquitted everybody except for defendant Song on attempted murder
• 26:58
  Speaker 6: and only as to one person. And so that means
• 27:01
  Speaker 6: that the jury likely did not find it compelling that
• 27:04
  Speaker 6: there was an attempt here from the broader conspiracy to
• 27:07
  Speaker 6: actually hurt or kill members of law enforcement, because of
• 27:11
  Speaker 6: course if they did, why wouldn't that be reasonably foreseeable
• 27:14
  Speaker 6: for purposes of Pinkerton liability? And so all of that,
• 27:17
  Speaker 6: and again the charges of this man defended Sanchez, who
• 27:20
  Speaker 6: wasn't there and wasn't involved in any planning, I think
• 27:22
  Speaker 6: speaks to the you know, the lack of plausibility and
• 27:26
  Speaker 6: some of the more nefarious strains that are being alleged
• 27:28
  Speaker 6: by the Department of Justice.
• 27:30
  Speaker 3: And so what's what's next here for these defendants?
• 27:34
  Speaker 4: Like it, are there constitutional arguments that can be made
• 27:38
  Speaker 4: on a on appeal?
• 27:39
  Speaker 3: Are they are they going to be made? Are they
• 27:42
  Speaker 3: free on bond?
• 27:42
  Speaker 6: Now?
• 27:43
  Speaker 3: Are they currently locked up?
• 27:44
  Speaker 2: Like?
• 27:44
  Speaker 3: What's the status?
• 27:47
  Speaker 6: Yeah, So there are a ton of legal issues that
• 27:49
  Speaker 6: were presented and rejected by the trial judge. I'll let
• 27:52
  Speaker 6: you know that the trial judge was a George Bush appointee,
• 27:55
  Speaker 6: and I believe the sendency either was helped with a
• 27:58
  Speaker 6: Trump appointee. So there are Republican judges sort of in
• 28:01
  Speaker 6: charge of Republican appointed judges in charge of some of
• 28:03
  Speaker 6: the proceedings here. But these issues sort of range from
• 28:07
  Speaker 6: issues regarding search and seizure that might have been might
• 28:11
  Speaker 6: be sort of tossed up for appeal, issues related to
• 28:13
  Speaker 6: the sufficiency of the evidence for some of these charges.
• 28:16
  Speaker 6: I mean, there are deep factual disputes about whether or
• 28:19
  Speaker 6: not you can charge any of these defendants, for example,
• 28:22
  Speaker 6: with rioting using the federal statute there, because there doesn't
• 28:26
  Speaker 6: appear to be any use of the federal instrumentality of
• 28:28
  Speaker 6: interstate commerce or any sort of crossing of state lines
• 28:31
  Speaker 6: that would invoke the federal statute, since all that was
• 28:33
  Speaker 6: sort of planned about or talk through these channels was
• 28:35
  Speaker 6: was not like riotous activity. So that's sort of going
• 28:39
  Speaker 6: to be up for appeal. The sufficiency of the evidence
• 28:41
  Speaker 6: regarding the material support. So there's a lot of, you know,
• 28:45
  Speaker 6: questions that are left for appeal and immediate issues that
• 28:47
  Speaker 6: the courts of appeals are going to have to deal with.
• 28:50
  Speaker 6: The thing is, though they're in the Fifth Circuit, which is,
• 28:52
  Speaker 6: you know, traditionally very conservative, very you know, Republican leaning.
• 28:56
  Speaker 6: And additionally, the trial judges are given a lot of
• 28:58
  Speaker 6: deference on like evidence issues, and they're given a lot
• 29:02
  Speaker 6: of difference also on some of these rulings, and so
• 29:04
  Speaker 6: plenty of issues to work with. And of course these
• 29:06
  Speaker 6: are federal convictions and so presumably if the next president
• 29:10
  Speaker 6: thought that these were two harsh sentences, I mean, there
• 29:12
  Speaker 6: is recent precedent for pardons and commutations of people involved
• 29:17
  Speaker 6: with conspiracies to overturn an election in the case of
• 29:20
  Speaker 6: January sixth, So those are kind of issues that are
• 29:23
  Speaker 6: still available and remedies available to the defendants and.
• 29:25
  Speaker 2: Peace go Finally, how does NSPM seven play into this
• 29:29
  Speaker 2: and what are the broader implications here in terms of
• 29:32
  Speaker 2: trying to you know, scare other activists and protesters out
• 29:37
  Speaker 2: of you know, out of participating in any sort of
• 29:40
  Speaker 2: descent against this government.
• 29:42
  Speaker 6: Yeah, politically, I think the administration is a huge win because,
• 29:45
  Speaker 6: as you know, that memorandum that you just referenced, one
• 29:48
  Speaker 6: of its goals is to try to like broaden the
• 29:50
  Speaker 6: scope of terrorism and to basically go after things that
• 29:55
  Speaker 6: maybe traditionally Americans would considered terrorism, and to sort of
• 29:59
  Speaker 6: increase the amount of surveillance over these kinds of groups,
• 30:01
  Speaker 6: especially in the aftermath of the killing of Charlie Kirk.
• 30:04
  Speaker 6: And so it's relevant because these I believe are the
• 30:07
  Speaker 6: first convictions of terrorism related charges for anyone involved with Antifa,
• 30:12
  Speaker 6: and so the administration is banding around as a win,
• 30:14
  Speaker 6: especially with the super harsh sentences, and so I think
• 30:16
  Speaker 6: this will encourage the administration to try these different theories.
• 30:20
  Speaker 6: One of the big complaints of the defendants in these
• 30:22
  Speaker 6: cases is that they relied a lot on this kind
• 30:24
  Speaker 6: of like innuendo and the association, and they shouldn't have
• 30:28
  Speaker 6: brought up the fact that they were related to ANTIFO,
• 30:30
  Speaker 6: or that they lingered too long on it. And so
• 30:33
  Speaker 6: the fear, perhaps from people who are against the administration,
• 30:36
  Speaker 6: is that this would encourage them to continue down this
• 30:39
  Speaker 6: road of the memorandum and identify members who are allegedly
• 30:43
  Speaker 6: part of these left wing groups and to ascribe to
• 30:46
  Speaker 6: them criminal culpability using doctrines like Pinkerton liability or some
• 30:49
  Speaker 6: of these also like conspiracy against rights cases that have
• 30:52
  Speaker 6: been sort of levied by the administration, but it is
• 30:55
  Speaker 6: a big sort of win in a series of political
• 30:58
  Speaker 6: or allegedly political prosecutions or sometimes they've had some failures recently.
• 31:01
  Speaker 2: Wow, unbelievable, pieas Go, thank you so much for your analysis.
• 31:05
  Speaker 2: I really appreciate it. Everybody, go subscribe to Piece go
• 31:08
  Speaker 2: at Peace Go Litty on YouTube. His channel is fantastic.
• 31:10
  Speaker 2: He does a great job breaking down all sorts of things,
• 31:12
  Speaker 2: legal cases but other things besides. Great to see a
• 31:14
  Speaker 2: piece go.
• 31:15
  Speaker 5: Thanks a lot.
• 31:19
  Speaker 2: So we are very fortunate this morning to be joined
• 31:21
  Speaker 2: by Cory Doctor. Oh you guys might be really familiar
• 31:23
  Speaker 2: with his work and shittification, which I confess I haven't
• 31:26
  Speaker 2: read but really should, but still seems to encapsulate a
• 31:28
  Speaker 2: lot of my experience of life just the word alone.
• 31:31
  Speaker 2: And he's out with a new book called The Reverse
• 31:33
  Speaker 2: Centaurs Guide to Life After AI that gives his view
• 31:37
  Speaker 2: of AI development. What's real, what's false? Is it a
• 31:40
  Speaker 2: hype machine? Is it a bubble? How should you approach it?
• 31:43
  Speaker 2: And there's a warning here to avoid being what he
• 31:46
  Speaker 2: describes as a reverse centaur, which is some industry lingo. Corey, welcome,
• 31:50
  Speaker 2: great to have you.
• 31:51
  Speaker 5: Hey, it's great to be on. Thank you for having me.
• 31:53
  Speaker 2: Yeah, of course. So I guess let's start there. What
• 31:55
  Speaker 2: is the reverse center? For people who are not.
• 31:56
  Speaker 5: In the know, Well, you know, this isn't the first
• 31:59
  Speaker 5: time automation and labor have come into conflict, and so
• 32:02
  Speaker 5: there's a pretty rich literature about how automation diffuses through workforces.
• 32:07
  Speaker 5: And one of the things that we know is that
• 32:09
  Speaker 5: when labor is in charge of how they use automation, broadly,
• 32:12
  Speaker 5: they tend to use it to increase the quality of
• 32:15
  Speaker 5: what they produce, Whereas when capital is in charge, they
• 32:19
  Speaker 5: broadly use it to increase the throughput. Because if you
• 32:21
  Speaker 5: acquire an asset and it's depreciating off your books, you
• 32:24
  Speaker 5: want to use it as much as possible. So this
• 32:26
  Speaker 5: is where we get centaurs and reverse centaurs and automation theory.
• 32:30
  Speaker 5: A centaur is someone who is directing a machine, you know,
• 32:33
  Speaker 5: like a human head on a machine body. So you know,
• 32:35
  Speaker 5: if you're riding a bicycle, you even look like a centaur.
• 32:37
  Speaker 5: About maybe you're using a spell checker or some other
• 32:40
  Speaker 5: tool that is allowing you to reduce your work better.
• 32:44
  Speaker 5: A reverse center is the reverse right. It's where you
• 32:47
  Speaker 5: aren't directing the way that a machine works. It's where
• 32:49
  Speaker 5: a machine is directing how you work. It's where you
• 32:52
  Speaker 5: are the human body that is supporting the machine brain.
• 32:56
  Speaker 5: And you know, you can think of things like Amazon
• 32:58
  Speaker 5: warehouse workers. Those are the most automated warehouses in America.
• 33:01
  Speaker 5: They also have three times the injury rate of other
• 33:03
  Speaker 5: warehouses in America. And that's because when you buy a
• 33:06
  Speaker 5: seven figure asset in the form of warehouse automation, you
• 33:10
  Speaker 5: want to maximize the value of it, and so you
• 33:12
  Speaker 5: work it at the limit of the speed of the system,
• 33:14
  Speaker 5: and the bottleneck of the speed of the system is
• 33:16
  Speaker 5: the person. And a person who works as fast and
• 33:19
  Speaker 5: as long as they can will eventually make a mistake,
• 33:21
  Speaker 5: which in a warehouse sometimes means you get impaled on
• 33:23
  Speaker 5: a forkliff.
• 33:26
  Speaker 3: So correct me if I'm wrong.
• 33:28
  Speaker 4: The book kind of either grew out of a lecture
• 33:32
  Speaker 4: that you gave back in twenty twenty five, or you
• 33:34
  Speaker 4: were already working in the book and gave the lecture,
• 33:38
  Speaker 4: and that lecture I remember, went viral to the time
• 33:40
  Speaker 4: a lot of people were reading it. Can you just
• 33:43
  Speaker 4: lay out for people, like what the thesis is of
• 33:46
  Speaker 4: that lecture, and sure, the broader point you're trying to
• 33:49
  Speaker 4: make in the book here.
• 33:51
  Speaker 5: Yeah, the story is even a little weirder than it
• 33:54
  Speaker 5: coming out of a lecture. The lecture and the story
• 33:57
  Speaker 5: and the book both came out of the fact that
• 33:59
  Speaker 5: I was sick of people demanding that I talked to
• 34:01
  Speaker 5: them about AI, and I thought like, if I just
• 34:03
  Speaker 5: compile all my thoughts in one place, I can be done,
• 34:06
  Speaker 5: not having realized that this would mean I would then
• 34:08
  Speaker 5: have to go and talk about AI. So the book
• 34:11
  Speaker 5: is really an inquiry into the sources of the bubble. Right,
• 34:14
  Speaker 5: Obviously there's some there there with the technology. You know,
• 34:17
  Speaker 5: speaking us a pretend computer scientists with an honorary degree
• 34:21
  Speaker 5: in CS, I have to say, like, as a software matter,
• 34:23
  Speaker 5: this is pretty cool stuff. But we spent one point
• 34:26
  Speaker 5: four trillion dollars on this technology so far as a species,
• 34:31
  Speaker 5: seven hundred billion of it in the last year, so
• 34:33
  Speaker 5: it's all time spending before that seven billion, another seven
• 34:36
  Speaker 5: hundred billion last year. Sam Altman says it's going to
• 34:39
  Speaker 5: be two trillion more next year. Why are we spending
• 34:42
  Speaker 5: this much money. It's not because it's making us money.
• 34:44
  Speaker 5: This is the money losing this thing the human race
• 34:46
  Speaker 5: has ever done. The gross revenues for the whole industry
• 34:49
  Speaker 5: worldwide or fifty billion dollars, which sounds like a lot
• 34:52
  Speaker 5: until you think about how much one point four trillion
• 34:54
  Speaker 5: is and you realize that they're depreciating these assets on
• 34:57
  Speaker 5: a five year schedule, but actually replacing them every two
• 34:59
  Speaker 5: to three years, which means that they have three years
• 35:01
  Speaker 5: of revenue of fifty billion dollars a year before they
• 35:04
  Speaker 5: have to buy another seven hundred billion dollars worth of stuff.
• 35:06
  Speaker 5: And that just doesn't pencil out. And the unit economics
• 35:10
  Speaker 5: are getting worse, not better. Every generation of this technology
• 35:12
  Speaker 5: loses more money. So where's it coming from? And the
• 35:16
  Speaker 5: thesis as well. On the one hand, you have a
• 35:18
  Speaker 5: lot of tech bubbles. This isn't the first one, and
• 35:20
  Speaker 5: the tech bubbles of this century have been driven by
• 35:25
  Speaker 5: the narrative need of companies that have saturated their markets
• 35:29
  Speaker 5: to convince Wall Street that they can keep growing. You know,
• 35:32
  Speaker 5: when Google attains a ninety percent search market share, that
• 35:35
  Speaker 5: sounds very good until you realize that this means that
• 35:38
  Speaker 5: Google is going to stop being a growth stock and
• 35:39
  Speaker 5: start being a mature stock, because you can't grow from
• 35:42
  Speaker 5: a ninety percent search market share, and a growth stock
• 35:44
  Speaker 5: is where a fraction of what a mature stock is
• 35:47
  Speaker 5: and growth stock can be used to grow. Right, you
• 35:50
  Speaker 5: can buy other companies with your highly liquid growth stock.
• 35:52
  Speaker 5: You can hire key personnel, and that's great because it's
• 35:58
  Speaker 5: much easier to get more of your own stock than
• 35:59
  Speaker 5: it is to get money. You make more stock by
• 36:02
  Speaker 5: typing zeros into a spreadsheet. You only get money from
• 36:05
  Speaker 5: like a creditor or a customer or an investor. The
• 36:09
  Speaker 5: problem is that when you stop growing, then you get
• 36:11
  Speaker 5: a massive haircut from the market. The net worth of
• 36:13
  Speaker 5: everyone who has been paid in stock plummets. The company
• 36:17
  Speaker 5: can't pull itself out of the death spiral by buying
• 36:20
  Speaker 5: more companies with stock, and so it's really bad. And
• 36:23
  Speaker 5: so companies have been telling us stories about how they're
• 36:26
  Speaker 5: going to grow into imaginary markets for a couple of decades.
• 36:29
  Speaker 5: Now metaverse and cryptocurrency, NFTs and dows web three and
• 36:34
  Speaker 5: so on, and now we get AI and so that's
• 36:38
  Speaker 5: the material foundation of it. I also, and maybe we
• 36:41
  Speaker 5: can talk about this next, explore the ideological basis. Why
• 36:44
  Speaker 5: are we spending one point four trillion on this, which
• 36:47
  Speaker 5: is so much more than we spent on NFTs and
• 36:50
  Speaker 5: cryptocurrency and web three in the metaverse.
• 36:52
  Speaker 2: Let's pause on the bubble question because there's true potent. Well,
• 36:55
  Speaker 2: I'm sure there's many potential types of bubbles, but for
• 36:57
  Speaker 2: our purposes, there's two potential types. The NFT metaverse type,
• 37:03
  Speaker 2: which is all hype and literally nothing there.
• 37:06
  Speaker 1: Right.
• 37:06
  Speaker 2: I agree with you that crypto is the same. So
• 37:08
  Speaker 2: there are those that would disagree. But you know, all
• 37:10
  Speaker 2: the blockchain technology whatever, No, it's a speculative asset bubble, etc.
• 37:14
  Speaker 2: So there's that type. And then there's also the dot
• 37:17
  Speaker 2: com bubble, which was a bubble which did pop. There
• 37:20
  Speaker 2: was a massive crash, huge implications, tons of companies go under.
• 37:23
  Speaker 2: But I think any reasonable person would say the Internet
• 37:25
  Speaker 2: itself was genuinely transformative. So which category do you put
• 37:29
  Speaker 2: the AI bubble in?
• 37:31
  Speaker 5: So I think that that's the wrong categorical cleavage. I
• 37:34
  Speaker 5: think that there's a slightly different categorical cleavage, which is
• 37:36
  Speaker 5: bubbles that leave something useful behind in bubbles that don't.
• 37:39
  Speaker 5: So some bubbles, all the companies fail and there's nothing
• 37:43
  Speaker 5: good left behind from it. In terms of a business,
• 37:45
  Speaker 5: there is no stock you could have bought beforehand that
• 37:47
  Speaker 5: would have been a goodbye. But then there's a lot
• 37:50
  Speaker 5: of stuff left behind that we used to start something else.
• 37:53
  Speaker 5: So you know, Enron left nothing behind. It was just
• 37:55
  Speaker 5: accounting fraud. WorldCom is accounting fraud. But the convincer was
• 37:58
  Speaker 5: they put thousands of miles of five in the ground
• 38:00
  Speaker 5: that's still there. So you know, my house in Burbank,
• 38:03
  Speaker 5: I've got two gigabit symmetrical fiber on old worldconfiber that
• 38:06
  Speaker 5: AT and T bought for pennies and lit up right.
• 38:08
  Speaker 5: And I think that the dot com bubble was one
• 38:10
  Speaker 5: of those. You had data centers, you had servers, you
• 38:13
  Speaker 5: had a trained workforce, all these humanities, undergraduates who were
• 38:17
  Speaker 5: like convinced to drop out of university and learn like
• 38:19
  Speaker 5: Python and Perl and HTML, who went on to build
• 38:21
  Speaker 5: Web two with all those cheap servers that were now
• 38:23
  Speaker 5: available for pennies on the dollar. You had all the
• 38:27
  Speaker 5: literacy in terms of how to use the web that
• 38:29
  Speaker 5: came out of spending all that money to promote it.
• 38:31
  Speaker 5: And AI definitely is that kind of bubble. It has
• 38:34
  Speaker 5: a lot of productive residue. There's going to be data centers,
• 38:37
  Speaker 5: there's going to be GPUs at ten cents on the dollar.
• 38:39
  Speaker 5: There's going to be skilled workers, there's going to be
• 38:41
  Speaker 5: open source models we can rehabilitate. But I don't know
• 38:44
  Speaker 5: if there's an Amazon there right Because the other difference
• 38:46
  Speaker 5: here is that the web had good unit economics. New
• 38:50
  Speaker 5: web users made the web more profitable. New Web uses
• 38:53
  Speaker 5: made the web more profitable. Subsequent generations of the Web
• 38:56
  Speaker 5: were more profitable than the previous ones. The opposite is
• 38:58
  Speaker 5: true of AI, right eat new customer loses money for
• 39:01
  Speaker 5: the company. The companies that use the customers that use
• 39:04
  Speaker 5: it the most cost them the most money. And every
• 39:06
  Speaker 5: generation is more money losing than the previous one. And
• 39:10
  Speaker 5: unlike say nineteen ninety nine or two thousand, when you
• 39:12
  Speaker 5: had Harvard Business Review and ber publishing these papers going
• 39:15
  Speaker 5: like our CEO is ready for workers who demand the web.
• 39:19
  Speaker 5: Now we have articles in those same outlets saying how
• 39:22
  Speaker 5: can CEOs force their workers to use AI? So I
• 39:25
  Speaker 5: think that's a very different kind of character to the bubble,
• 39:28
  Speaker 5: although on this material level of like will there be
• 39:31
  Speaker 5: something left over? I think the answer is yes.
• 39:33
  Speaker 2: It is like the dot com bubble in that, don't
• 39:35
  Speaker 2: they Chinese models have better per unit costs?
• 39:40
  Speaker 5: Yes, but they're doing something different. So if you know
• 39:42
  Speaker 5: with the Chinese models that you host yourself right, that
• 39:44
  Speaker 5: are open source, that you host yourself, and that aren't
• 39:46
  Speaker 5: trying to do this frontier compute where you're basically selling
• 39:50
  Speaker 5: hundred dollar bills for a dollar apiece, have better unit economics.
• 39:54
  Speaker 5: We have such opacity in how these AI companies work here,
• 39:58
  Speaker 5: the ones that do the so called frontier models, it's
• 40:00
  Speaker 5: really hard to know exactly what's going on. Ed Zitron
• 40:03
  Speaker 5: published some leak financials from open ai recently that makes
• 40:06
  Speaker 5: it pretty clear that like their claims that they can
• 40:09
  Speaker 5: do what's called inference. So the inference part is like
• 40:11
  Speaker 5: you type a question into chat GBT and it answers it,
• 40:14
  Speaker 5: and obviously that's like the operating expense. The training is
• 40:17
  Speaker 5: the is the capital expense. But the operating expense they say, oh,
• 40:21
  Speaker 5: it's really low, But it looks like what they're doing
• 40:24
  Speaker 5: is they're hiding a lot of their inference and their
• 40:26
  Speaker 5: marketing budgets. So what they're saying is like, some of
• 40:28
  Speaker 5: those hundred dollar bills we're giving away for a dollar
• 40:31
  Speaker 5: are not inference, even though that's what we're spending ninety
• 40:34
  Speaker 5: nine dollars for for the dollar and revenue, that's actually
• 40:37
  Speaker 5: a marketing expense where we're suckering in new customers by
• 40:40
  Speaker 5: selling them ninety one hundred dollar bills for a buck.
• 40:42
  Speaker 5: So if we don't count that as an infest expenditure,
• 40:45
  Speaker 5: we count that as a marketing expenditure. And so you
• 40:47
  Speaker 5: see these marketing budgets that you know, as Zitron says,
• 40:50
  Speaker 5: would make them like significantly larger than say Coca Cola
• 40:54
  Speaker 5: as a marketing venture, and we just don't see the
• 40:59
  Speaker 5: ad agency with the personnel that would absorb that marketing capital.
• 41:03
  Speaker 5: So it really looks like there's a shell game there.
• 41:05
  Speaker 2: Okay, interesting on the you know, the technological front. So
• 41:09
  Speaker 2: the theory both from people inside of these companies and
• 41:13
  Speaker 2: also some of the alarms that are being sounded both
• 41:16
  Speaker 2: from executives who have left these companies, but also from
• 41:19
  Speaker 2: you know, researchers, from professors, from experts in the field,
• 41:23
  Speaker 2: some of the godfathers of AI. There's a variety of
• 41:26
  Speaker 2: the warnings, but let's take the most material one first,
• 41:28
  Speaker 2: that it's going to significantly displace some portion of the workforce. Now,
• 41:33
  Speaker 2: you will get some people within the industry that say
• 41:35
  Speaker 2: the entire The goal is to replace all of human labor. Okay,
• 41:38
  Speaker 2: that is the goal that they state, But you also
• 41:40
  Speaker 2: have people who are not interested in, you know, overhyping
• 41:43
  Speaker 2: the technology, who are genuinely trying to sound the alarm.
• 41:46
  Speaker 2: And when I look at things like the fact that
• 41:47
  Speaker 2: you know, you mentioned you have a computer science background
• 41:50
  Speaker 2: for people who don't know, you're also a sci fi writer,
• 41:51
  Speaker 2: So you think about these, you know, these sorts of
• 41:53
  Speaker 2: potential scenarios. The fact that it has already developed to
• 41:57
  Speaker 2: the point where most coders are not actually coding anymore.
• 42:01
  Speaker 2: They're overseeing these machines that are doing the coding. They're
• 42:05
  Speaker 2: doing sort of the creative process. They're checking on the
• 42:07
  Speaker 2: other end and making sure that all works properly. And
• 42:10
  Speaker 2: that the market for students coming out of college to
• 42:13
  Speaker 2: be coders has sort of fallen off a cliff. You know,
• 42:16
  Speaker 2: it used to five seconds ago. The whole mantra was
• 42:18
  Speaker 2: you need to learn to code, and then you've got
• 42:19
  Speaker 2: your golden ticket into a high paying career. That is
• 42:23
  Speaker 2: just completely gone at this point. So doesn't that indicate
• 42:27
  Speaker 2: that some of the labor disruption promise has already come
• 42:30
  Speaker 2: to fruition and there's no reason to expect that the
• 42:32
  Speaker 2: technology won't continue to advance and displace more and more
• 42:36
  Speaker 2: people in more industries.
• 42:37
  Speaker 5: So I think now you're getting to some of the
• 42:39
  Speaker 5: ideological questions about why we would spend one point four
• 42:42
  Speaker 5: trillion dollars, like why would capital allocators be willing to
• 42:45
  Speaker 5: spend so much money on this? And I think that
• 42:48
  Speaker 5: you don't have to believe that this can do people's
• 42:51
  Speaker 5: jobs in order to invest in it. You have to
• 42:53
  Speaker 5: believe that salespeople can convince employers that it can do
• 42:58
  Speaker 5: people's jobs. You're pushing on an open door. There is
• 43:02
  Speaker 5: again a long history of employers being extremely interested in
• 43:06
  Speaker 5: replacing mouthy workers who tell them that their ideas are
• 43:09
  Speaker 5: dumb and who do things that they don't know how
• 43:12
  Speaker 5: to do with software that just does whatever it's told.
• 43:16
  Speaker 5: I think if you're a boss, there's a certain haunting
• 43:19
  Speaker 5: knowledge that, although you fancy that you're in the driver's seat,
• 43:22
  Speaker 5: you know that if you don't show up for work,
• 43:23
  Speaker 5: nothing happens. Whereas if the workers don't show up for work,
• 43:26
  Speaker 5: the whole shop grinds to a halt. And so you
• 43:29
  Speaker 5: wonder if maybe you're in the backseat with a Fisher
• 43:31
  Speaker 5: price steering wheel, and that AI sales call is a
• 43:34
  Speaker 5: promise to wire the toy steering wheel into the car's
• 43:36
  Speaker 5: drive trains. So you have an amazing idea, you tell
• 43:39
  Speaker 5: the chatbot to make it, and the chatbot does not
• 43:42
  Speaker 5: tell you it's illegal. It doesn't tell you're going to
• 43:44
  Speaker 5: hurt a lot of people. It doesn't tell you that
• 43:46
  Speaker 5: algae blossom in a pool that you've painted dark blue.
• 43:50
  Speaker 5: It doesn't. It doesn't tell you that it won't scale,
• 43:52
  Speaker 5: it doesn't tell you that you're going to get hacked.
• 43:54
  Speaker 5: It just does it. And then you know, if you
• 43:56
  Speaker 5: fired nine tenths of your workers, the ones who used
• 43:58
  Speaker 5: to tell you off that and be so demoralizing to you,
• 44:02
  Speaker 5: those guys are so scared of losing their jobs that
• 44:05
  Speaker 5: you can get them to fix the commas and polish
• 44:08
  Speaker 5: it up and keep it running and go and skim
• 44:10
  Speaker 5: the pond scum out of the pool every day. And
• 44:13
  Speaker 5: so I think that that's the sales pitch. When you
• 44:16
  Speaker 5: described how coders use it, there was a little bit
• 44:19
  Speaker 5: of conflation there that I think is characteristic of the industry.
• 44:22
  Speaker 5: So you said, well, there are some workers who say
• 44:24
  Speaker 5: that they use AI. They do the creative part. And
• 44:26
  Speaker 5: then some of the utility functions are done by the AI,
• 44:29
  Speaker 5: and there are other workers who mark the AI's homework.
• 44:32
  Speaker 5: And so what you're saying is you have some centaurs
• 44:34
  Speaker 5: and some reversal, yes, and where workers are in charge
• 44:38
  Speaker 5: of their working conditions, which is the thing that tech
• 44:41
  Speaker 5: bosses don't like about their tech workers. Now, tech workers
• 44:44
  Speaker 5: are historically they've been in short supply. They're very productive
• 44:47
  Speaker 5: and valuable. When Silicon Valley firms hired an extra engineer,
• 44:51
  Speaker 5: they anticipated like an extra million dollars on their bottom line.
• 44:54
  Speaker 5: And so that's why, you know, Silicon Valley engineers got
• 44:57
  Speaker 5: like free kombucha and massages and you know, someone who'd
• 45:00
  Speaker 5: freeze their eggs for them so they could work through
• 45:02
  Speaker 5: their fertile years. It wasn't because as a class, Silicon
• 45:05
  Speaker 5: Valley bosses like treating workers well. Otherwise Amazon warehouse workers
• 45:08
  Speaker 5: wouldn't be peeing in bottles. Was because there are ten
• 45:10
  Speaker 5: bosses at the factor gates who wanted to give them
• 45:12
  Speaker 5: a job. And AI doesn't need to do the programmer's job.
• 45:17
  Speaker 5: It just needs to allow them to discipline the labor
• 45:21
  Speaker 5: force so that you don't get Google walkouts, saying we're
• 45:24
  Speaker 5: not going to make tech for drones so that you
• 45:28
  Speaker 5: don't get mouthy workers who demand higher wages who hop
• 45:32
  Speaker 5: from employer to employer. You remember that the Silicon Valley
• 45:35
  Speaker 5: firms about ten years ago were settled a big criminal
• 45:39
  Speaker 5: case where they had gotten together and done a no
• 45:42
  Speaker 5: poaching agreement, saying they wouldn't give raises to each other's
• 45:45
  Speaker 5: talent to bring them over because their wages were very
• 45:48
  Speaker 5: high and they had a lot of worker power. And
• 45:51
  Speaker 5: so an AI that is controlled by a skilled worker
• 45:55
  Speaker 5: doesn't solve that problem for bosses. And the AI that
• 46:00
  Speaker 5: that is being used in ways that are really impressive
• 46:04
  Speaker 5: for software development, that AI is the AI that doesn't
• 46:07
  Speaker 5: fulfill that goal. The AI that you know poops out
• 46:11
  Speaker 5: code faster than any human being could possibly review it.
• 46:14
  Speaker 5: That is then, you know, has its homework marked by
• 46:16
  Speaker 5: coders who are terrified they're going to lose their job
• 46:18
  Speaker 5: and are working as sergate. Brinn says a six hour week,
• 46:22
  Speaker 5: which he says is the target that everyone working with
• 46:24
  Speaker 5: an AI should be shooting for. Those workers do fulfill
• 46:29
  Speaker 5: the destiny. But those are the workers saying, you know,
• 46:31
  Speaker 5: I'm a software developer in aerospace, we're using AI to
• 46:35
  Speaker 5: write code. Never get on an airplane again, because we're
• 46:38
  Speaker 5: producing the most ghastly TechEd at scale. So you know,
• 46:41
  Speaker 5: until you can like cleave apart these subtle issues, you
• 46:44
  Speaker 5: can't know where AI is affecting the labor market, how
• 46:48
  Speaker 5: it's affecting the labor market, and how that relates to
• 46:51
  Speaker 5: the investor story.
• 46:53
  Speaker 4: Well on the question of hype, and correctly if I'm wrong,
• 46:57
  Speaker 4: but you seem to view this know what's coming from
• 47:01
  Speaker 4: the AI companies and the CEOs, you know as at
• 47:05
  Speaker 4: least primarily hype that it would be that it'll the
• 47:08
  Speaker 4: popping of the bubble, will leave beside it behind something.
• 47:10
  Speaker 3: But it's mostly hype.
• 47:12
  Speaker 4: And then if you can have a salesman who can
• 47:15
  Speaker 4: go to a company and say, hey, this will allow
• 47:17
  Speaker 4: you to replace your mouth workers, that that's push going
• 47:19
  Speaker 4: open door. But that is that isn't of course, a
• 47:23
  Speaker 4: long term business plan like that. It has to actually
• 47:27
  Speaker 4: has to actually work like so it's at some point
• 47:32
  Speaker 4: if if it's not, you're going to start to see
• 47:34
  Speaker 4: some dialing back, which you're you're starting to see a
• 47:37
  Speaker 4: little bit of that from companies now which you're finding
• 47:41
  Speaker 4: that actually you know, they're not able to completely replace
• 47:45
  Speaker 4: people at this point. So much of this seems to
• 47:47
  Speaker 4: be organized around kind of people's thoughts and motivations and
• 47:51
  Speaker 4: ideas rather than what's actually happening.
• 47:54
  Speaker 3: We can put up faight that what this is kind
• 47:56
  Speaker 3: of a.
• 47:57
  Speaker 4: Counter argument being made by one one of the one
• 48:01
  Speaker 4: of the AI folks is I think this is an
• 48:03
  Speaker 4: open AI person.
• 48:04
  Speaker 3: I can't remember it, he says.
• 48:06
  Speaker 4: He says, so many intelligent people still cling to the
• 48:09
  Speaker 4: idea that everything the AI labs do is downstream of IPO,
• 48:14
  Speaker 4: hope or whatever. Some kinds of intelligencia are uniquely unsuited
• 48:18
  Speaker 4: for the moment because there Lens is quote strategic skepticism,
• 48:21
  Speaker 4: where normal people can employ magical thinking. And the latter
• 48:24
  Speaker 4: part of that what he means by the magical thinking
• 48:28
  Speaker 4: is like is observing the AI mechanistically and saying, okay,
• 48:33
  Speaker 4: we have we have some ideas about how this is
• 48:36
  Speaker 4: working and why it's working, but we don't we don't
• 48:40
  Speaker 4: actually completely understand it, and there is something going on
• 48:43
  Speaker 4: here that is currently beyond our understanding of it.
• 48:49
  Speaker 3: Yet it is doing a lot of these very interesting things.
• 48:51
  Speaker 4: It's solving math problems that you know, humanity has not
• 48:54
  Speaker 4: been able to solve, which puts it beyond just kind
• 48:58
  Speaker 4: of a guessing game robot like a very significant thing.
• 49:01
  Speaker 3: So, while we're in the.
• 49:03
  Speaker 4: Realm of just ideas and motivations and everything, what do
• 49:06
  Speaker 4: you think of this challenge to kind of people like yourself,
• 49:11
  Speaker 4: people like me in some respects that we are uniquely
• 49:16
  Speaker 4: unsuited for this moment because our constitutions are constantly skeptical.
• 49:22
  Speaker 4: But at some point you're going to run up against
• 49:24
  Speaker 4: something that is real. And how are we going to
• 49:27
  Speaker 4: tell the difference? If we're constantly looking for the fakeness
• 49:33
  Speaker 4: behind the frauds that run these companies, what if actually
• 49:37
  Speaker 4: underneath it there are these computer scientists who actually have
• 49:41
  Speaker 4: developed something.
• 49:43
  Speaker 5: Well, I think it's a very weird argument that being
• 49:46
  Speaker 5: discerning makes it hard for you to find the truth,
• 49:49
  Speaker 5: you know, approaching things and saying well, that's an extraordinary claim.
• 49:52
  Speaker 5: I don't need any extraordinary evidence because the claim itself
• 49:55
  Speaker 5: is so interesting. I think is a very weird way
• 49:58
  Speaker 5: to do truth seeking. And I think if you want
• 50:01
  Speaker 5: to talk about magical thinking, the idea that shoveling words
• 50:03
  Speaker 5: into the word guessing program makes it wake up and
• 50:05
  Speaker 5: turn into God is extremely magical thinking. I always say,
• 50:10
  Speaker 5: it's like believing that if you keep reading horses to
• 50:12
  Speaker 5: run faster, eventually one of them gives birth to a locomotive. Right.
• 50:16
  Speaker 5: The thing that we don't know what consciousness is. We
• 50:20
  Speaker 5: have people who make AI who say things like this
• 50:23
  Speaker 5: is fifty one percent of the way to consciousness. And
• 50:26
  Speaker 5: when you challenge them and you say you don't know
• 50:27
  Speaker 5: what consciousness is, they say, well, how do you know
• 50:30
  Speaker 5: this is in consciousness if you don't know what consciousness is,
• 50:33
  Speaker 5: which is again a very strange little rhetorical slight. You know,
• 50:37
  Speaker 5: the last time someone said that to me, I said,
• 50:38
  Speaker 5: I don't know what the weather in Paris is and
• 50:40
  Speaker 5: I don't know what the weather in Lisbon is. That
• 50:42
  Speaker 5: doesn't mean they have the same weather. The fact that
• 50:44
  Speaker 5: two things that we don't understand exists doesn't mean that
• 50:46
  Speaker 5: they're the same thing. And while we don't know exactly
• 50:50
  Speaker 5: how AI works, and while AI does exceed many of
• 50:54
  Speaker 5: the technical guesses that we've made about how making small
• 50:59
  Speaker 5: changes in the techechniques that we apply to machine learning
• 51:02
  Speaker 5: would produce outputs that would be commensurately small and in fact,
• 51:06
  Speaker 5: and instead they're larger, which is always exciting. You know.
• 51:09
  Speaker 5: The formal term for that is a breakthrough. Breakthroughs are
• 51:12
  Speaker 5: cool and interesting that even though that's true, the world
• 51:17
  Speaker 5: is full of systems we don't fully understand and whose
• 51:19
  Speaker 5: inputs and outputs we don't control precisely, that are instead
• 51:24
  Speaker 5: in a kind of brittle equilibrium that makes all kinds
• 51:27
  Speaker 5: of sways and changes all the time, including just you know,
• 51:31
  Speaker 5: firms and even small, relatively well understood markets do things
• 51:37
  Speaker 5: that are chaotic and hard to anticipate. So I think
• 51:40
  Speaker 5: saying oh, well, you know, you're uniquely unsuited because you've
• 51:42
  Speaker 5: never dealt with the thing we don't fully understand, means
• 51:46
  Speaker 5: that you fail to grasp how many things we don't
• 51:48
  Speaker 5: fully understand, which is most of them.
• 51:51
  Speaker 4: So go ahead, Ryan, One I gets to counter arguments
• 51:57
  Speaker 4: that you hear from defenders of the virtues of the
• 52:01
  Speaker 4: or the power of the technology.
• 52:03
  Speaker 3: Are that calling it calling it like a word guessing game?
• 52:06
  Speaker 2: Is?
• 52:07
  Speaker 4: You know, minimizes it to a to a degree that
• 52:11
  Speaker 4: just misunderstands what's going on here? And one and the
• 52:14
  Speaker 4: two examples they often give are you know, a couple
• 52:17
  Speaker 4: that has been together for decades that is able to
• 52:20
  Speaker 4: complete each other's sentences. It's not just a kind of
• 52:23
  Speaker 4: random mathematical guessing game that they're doing. It's like they
• 52:26
  Speaker 4: actually deeply understand each other. And it's not they're not
• 52:29
  Speaker 4: just predicting from the past. There's there's something, there's reasoning
• 52:33
  Speaker 4: and understanding behind that. And then the other example they
• 52:35
  Speaker 4: give is, let's say you've you've read a mystery novel
• 52:39
  Speaker 4: or a sci fi novel, but it missed. Let's say
• 52:41
  Speaker 4: it's a mystery novel and in the last page they say, uh,
• 52:45
  Speaker 4: and the detective says, I understand. I figured out who
• 52:48
  Speaker 4: the killer is, and the killer is and then it stops.
• 52:51
  Speaker 4: And then the reader, if you ask the reader, what's
• 52:55
  Speaker 4: the word that's going to come after this, the reader
• 52:58
  Speaker 4: who can accurately guess the next word can do so
• 53:02
  Speaker 4: because of reasoning and understanding, like they understood the book,
• 53:06
  Speaker 4: they understood the story, and so it's not just mathematical
• 53:10
  Speaker 4: computations and guessing.
• 53:12
  Speaker 5: And so that's true when people I'm sorry, go ahead,
• 53:15
  Speaker 5: I cut you an.
• 53:17
  Speaker 4: So they're saying, so there is some level of reasoning
• 53:20
  Speaker 4: and understanding going on here that we don't quite understand
• 53:24
  Speaker 4: ourselves yet. How it mechanistically comes together, but it's there's
• 53:29
  Speaker 4: something there.
• 53:30
  Speaker 5: So it is true that when people do it, it's
• 53:32
  Speaker 5: because we're doing a thing called understanding and reasoning, which
• 53:35
  Speaker 5: we only have a cursory understanding of what's going on,
• 53:42
  Speaker 5: both mechanically within the brain and then you know, sort
• 53:46
  Speaker 5: of chemically and so on. Cognitively, what's going on is
• 53:48
  Speaker 5: not perfectly understood, but we actually have a much better
• 53:52
  Speaker 5: idea of what the computer is doing. And if we're
• 53:53
  Speaker 5: discovering something about what computers are doing, it's that we
• 53:57
  Speaker 5: are discovering that word guessing is actually very very power power,
• 54:01
  Speaker 5: which does not make it reasoning. So it turns out
• 54:04
  Speaker 5: that there are like linguistic correlates latent in the structure
• 54:08
  Speaker 5: of a mystery novel that make it possible to if
• 54:11
  Speaker 5: you count all the words, which is like, so the
• 54:14
  Speaker 5: idea that counting words and then guessing based on the
• 54:17
  Speaker 5: count is not what they're doing, is just technically wrong.
• 54:19
  Speaker 5: That is, in fact what they're doing. That's what model
• 54:22
  Speaker 5: training is, right. You're measuring the frequency distribution of words
• 54:26
  Speaker 5: in their relationship to other words and phrases. That's what
• 54:28
  Speaker 5: it is now saying that's all it is doesn't mean
• 54:32
  Speaker 5: that it's trivial. It means that there's a lot of
• 54:34
  Speaker 5: stuff that you can figure out by making those kinds
• 54:37
  Speaker 5: of guesses. But those guesses are not the same as comprehension,
• 54:40
  Speaker 5: which is where we get this funny term that we
• 54:42
  Speaker 5: use when we describe the defects in AI products, which
• 54:45
  Speaker 5: is hallucination. It's not a hallucination. It's just a limit
• 54:49
  Speaker 5: to statistical guessing. Now, one of the things that AI
• 54:53
  Speaker 5: did for us, this current generation of AI, is it
• 54:56
  Speaker 5: showed us that statistical guessing is more powerful than we
• 54:58
  Speaker 5: thought it would be when you gale it up. Yeah,
• 55:02
  Speaker 5: the returns to scale from adding more training data and
• 55:05
  Speaker 5: measuring more and more subtle relationships are larger than we
• 55:10
  Speaker 5: thought they'd be. That is a legitimately really interesting thing
• 55:14
  Speaker 5: to discover about computer science and about frankly, the nature
• 55:18
  Speaker 5: of reality. And it's not the same thing as reasoning.
• 55:21
  Speaker 5: Otherwise these models wouldn't make the characteristic errors that you
• 55:26
  Speaker 5: get when you're not reasoning but just guessing guessing words.
• 55:29
  Speaker 2: Right, So I guess some of what I would say
• 55:32
  Speaker 2: to that is is there any limit though to that
• 55:36
  Speaker 2: power of scale? Because there have been a lot of
• 55:38
  Speaker 2: predictions of Oh, they're going to hit a wall, they're
• 55:40
  Speaker 2: not going to get better, they're not going to improve
• 55:42
  Speaker 2: much from here, and that hasn't been the case. They've
• 55:44
  Speaker 2: continued to improve an open ay model solve this eighty
• 55:49
  Speaker 2: year old mathematics problem that humans had been stumped on.
• 55:52
  Speaker 2: There is a situation that unfolder, we could put F
• 55:54
  Speaker 2: seven up on the screen. I'm sure you probably followed this.
• 55:57
  Speaker 2: There's a whole controversy over whether a CHAT wrote this
• 56:01
  Speaker 2: prize winning story, you know, for this famous Commonwealth Short
• 56:06
  Speaker 2: Story prize. There were a lot of people who looked
• 56:08
  Speaker 2: at this and oh, this is AI writing. They dispute
• 56:11
  Speaker 2: that it was. You know, it seems like the analysis
• 56:13
  Speaker 2: at this point is that it wasn't actually AI writing.
• 56:15
  Speaker 2: But I think it indicates how advanced the AI, even
• 56:20
  Speaker 2: in creative endeavors that require, you know, the sorts of
• 56:23
  Speaker 2: things that you would want to imagine humans were sort
• 56:25
  Speaker 2: of uniquely good at that AI has gotten to that
• 56:27
  Speaker 2: place where it's indistinguishable. I've seen other analyzes where experts
• 56:30
  Speaker 2: are judging short stories and they're rating the you know,
• 56:33
  Speaker 2: the AI stories just as highly er if not higher,
• 56:35
  Speaker 2: than some of the human creative product. So you know,
• 56:39
  Speaker 2: if these laws of scale that just adding more and
• 56:42
  Speaker 2: more and more genuinely do produce these sorts of profound improvements,
• 56:47
  Speaker 2: what is ultimately the limit to that?
• 56:49
  Speaker 5: So the scale is breaking in the sense that they is.
• 56:53
  Speaker 5: It used to be that linear inputs to scale produce
• 56:55
  Speaker 5: linear outputs to scale or even better than linear. So
• 56:58
  Speaker 5: you spent an extra dollar, you got an extra units
• 57:00
  Speaker 5: worth of goodness out the other side. Whatever we're using
• 57:03
  Speaker 5: to benchmark that that number is going down the amount
• 57:05
  Speaker 5: that you get out, which means that we're having to
• 57:07
  Speaker 5: put much more. And that's why each model, you know,
• 57:09
  Speaker 5: each new foundation model is less profitable than the previous ones,
• 57:13
  Speaker 5: is that they are hitting limits to scale.
• 57:16
  Speaker 2: You know.
• 57:17
  Speaker 5: I have heard people say that they have seen AI
• 57:19
  Speaker 5: art that they can't distinguish from human art of quality.
• 57:24
  Speaker 5: And I read that short story and it was ghastly.
• 57:27
  Speaker 5: I mean, it's a it's a bad short story. I
• 57:29
  Speaker 5: think that one of the res you know, and it's
• 57:32
  Speaker 5: not the first time that there's been a literary prize
• 57:35
  Speaker 5: awarded to a piece of fiction that I looked at
• 57:38
  Speaker 5: and was like, this is unreadable, terrible stuff. And maybe
• 57:41
  Speaker 5: that's my bias because I'm a science fiction writer. We
• 57:44
  Speaker 5: write pulp literature. And you know, within the science fiction field,
• 57:48
  Speaker 5: which is one of the last feels that actually pays
• 57:50
  Speaker 5: for short fiction, science fiction editors are being drowned in
• 57:54
  Speaker 5: AI stories and they don't have any trouble discerning them
• 57:57
  Speaker 5: because they are each of them less publishable, publishable than
• 58:00
  Speaker 5: the last. They are dreadful, dreadful work. You know, I'm
• 58:05
  Speaker 5: not a mathematician. I'm the son of a mathematician. I
• 58:07
  Speaker 5: don't know enough about math to assess the problems that
• 58:10
  Speaker 5: have been solved. It seems really impressive. And you know,
• 58:13
  Speaker 5: the fact that mathematicians have figured out how to use
• 58:16
  Speaker 5: computers to solve problems that they couldn't solve before is
• 58:19
  Speaker 5: both wonderful and normal. Mathematicians have solved new problems with
• 58:23
  Speaker 5: using computers for a really long time. And you know,
• 58:26
  Speaker 5: a mathematician of my acquaintance was talking about iterating through
• 58:30
  Speaker 5: several different ways of approaching a proof, which would be
• 58:35
  Speaker 5: labor intensive to do personally, in which you can automate
• 58:38
  Speaker 5: through a computer. That's great. You know, look, there's a
• 58:40
  Speaker 5: lot of things that this software does that I think
• 58:43
  Speaker 5: in the absence of a bubble we would just call
• 58:45
  Speaker 5: a plug in, and there have been some really important
• 58:47
  Speaker 5: plugins for the tools I use over my life. Word
• 58:51
  Speaker 5: processors have really made some leaps and bounds. My first
• 58:53
  Speaker 5: word processor was a program printed in a magazine. I
• 58:56
  Speaker 5: bought it a newsstand and typed into an Apple two plus.
• 59:00
  Speaker 5: In the years since, where processors have gotten all kinds
• 59:02
  Speaker 5: of features, some of which are very valuable to me,
• 59:04
  Speaker 5: some of which I think are very frivolous, some which
• 59:07
  Speaker 5: I think when other people use them it makes their
• 59:08
  Speaker 5: work worse. I've never been in a circumstance where I said, well,
• 59:11
  Speaker 5: this plugin is so interesting, let's bet the economy on it.
• 59:14
  Speaker 5: And also it's probably time to feed the writers into
• 59:16
  Speaker 5: a wood chair.
• 59:16
  Speaker 2: Yeah, Corey, My last question for you is do you
• 59:21
  Speaker 2: worry at all that if you're wrong, that if the
• 59:24
  Speaker 2: technology is maybe not as you know, we're going to
• 59:27
  Speaker 2: replace all human beings, but if it is significantly you know,
• 59:30
  Speaker 2: more powerful, continues to improve and there is some large
• 59:33
  Speaker 2: chunk of the labor force that is displaced. Do you
• 59:36
  Speaker 2: worry that if you're wrong and downplaying its capabilities now,
• 59:40
  Speaker 2: that you'll suck some of the energy out of the
• 59:42
  Speaker 2: capability of people to organize in ways that are needed
• 59:45
  Speaker 2: to deal with that you know, very scary potential future.
• 59:49
  Speaker 5: I actually think that, you know, organizing about AI is
• 59:54
  Speaker 5: a useful endeavor and that we should be organizing about it.
• 59:57
  Speaker 5: I think the local movements that are talking about data
• 1:00:00
  Speaker 5: center justice. You know, no city council ever said, oh,
• 1:00:02
  Speaker 5: we approve this planning permission and exercised eminine domain on
• 1:00:06
  Speaker 5: your farm without talking to you about it, because we
• 1:00:08
  Speaker 5: knew that you'd be really pleasantly surprised and we didn't
• 1:00:10
  Speaker 5: want to wreck it. And so the fact that we're
• 1:00:12
  Speaker 5: getting these local movements that are building up this capacity
• 1:00:16
  Speaker 5: to demand equitable treatment and to demand code determination with technology,
• 1:00:20
  Speaker 5: I think is great. I mean, if it were the
• 1:00:24
  Speaker 5: case that tomorrow AI could do a lot of human
• 1:00:27
  Speaker 5: jobs that currently are being done by humans, a functional
• 1:00:32
  Speaker 5: economy would find work for those people, because we have
• 1:00:34
  Speaker 5: full employment. For the next three hundred years, We're gonna
• 1:00:37
  Speaker 5: have to move like coastal cities ten kilometers inland. We're
• 1:00:40
  Speaker 5: gonna have to deal with hundreds of millions of climate refugees.
• 1:00:43
  Speaker 5: We're gonna have more zoonotic plagues as we get more
• 1:00:45
  Speaker 5: habitat loss. Like you me and everyone we know and
• 1:00:48
  Speaker 5: everyone who will live for the next three centuries is
• 1:00:51
  Speaker 5: going to have their hands full. Like we don't have
• 1:00:53
  Speaker 5: a technological unemployment problem on our hands. If we have
• 1:00:56
  Speaker 5: an unemployment problem, it's because we have a social and
• 1:00:58
  Speaker 5: economic problem that is exacerbating our environmental problem.
• 1:01:03
  Speaker 2: Well, that sounds like a grim future, just not one
• 1:01:05
  Speaker 2: where the central problem is unemployment. Corey, thank you so much.
• 1:01:09
  Speaker 2: I really appreciate your time this morning, and I think
• 1:01:12
  Speaker 2: your ideas are really fascinating. Appreciate you spending the time
• 1:01:14
  Speaker 2: to engage with this on them. The book is The
• 1:01:16
  Speaker 2: Reverse Centaur's Guide to Life After AI, and I recommend
• 1:01:19
  Speaker 2: everybody pick it up.
• 1:01:21
  Speaker 5: And if you want to get involved on these issues,
• 1:01:23
  Speaker 5: visit EFF dot org. The Electronic Frontier Foundation has been
• 1:01:25
  Speaker 5: doing this work for thirty five years and I've been
• 1:01:28
  Speaker 5: doing it with them for twenty five years.
• 1:01:29
  Speaker 2: Amazing. Thank you for that plug. That's an interesting conversation, Ran,
• 1:01:33
  Speaker 2: I like that one.
• 1:01:35
  Speaker 4: Yeah.
• 1:01:36
  Speaker 3: I think your question is an important one. What if
• 1:01:38
  Speaker 3: he's wrong?
• 1:01:39
  Speaker 4: And to me, there's sort of a Pascal's Wager element
• 1:01:43
  Speaker 4: to it. Maybe it's all hype. If it is, then
• 1:01:47
  Speaker 4: it's another bubble that comes and goes. But if it's not,
• 1:01:51
  Speaker 4: you got to be planning for it now. Yeah, we
• 1:01:55
  Speaker 4: have to be thinking and organizing about what the world
• 1:01:58
  Speaker 4: looks like in the event that it's not. And I
• 1:02:00
  Speaker 4: don't think it's I don't think it's hype. So I'm
• 1:02:03
  Speaker 4: in the minority, and you and I think we're both
• 1:02:05
  Speaker 4: in the minority on this on the left.
• 1:02:08
  Speaker 3: But I think it's something that you know, we're going
• 1:02:11
  Speaker 3: to hopefully live long enough to find out.
• 1:02:13
  Speaker 2: Yeah. Well, the other thing is, let's say it is hype.
• 1:02:16
  Speaker 2: The very fact that you have all of these oligarchs
• 1:02:18
  Speaker 2: who are like, we need to change the social contract.
• 1:02:21
  Speaker 2: That also seems like an opening because I would like
• 1:02:23
  Speaker 2: to change the social I don't like the social contract
• 1:02:25
  Speaker 2: that we have now. It leaves many billions of people
• 1:02:28
  Speaker 2: needlessly and miserated. So if they're giving us that opening,
• 1:02:31
  Speaker 2: even if it is to hype their you know, shitty products,
• 1:02:34
  Speaker 2: which I agree with you, I don't think that that is.
• 1:02:37
  Speaker 2: I do think that there's a lot of there there
• 1:02:38
  Speaker 2: for the technology. I do think it's going to be
• 1:02:40
  Speaker 2: genuinely transformational. I also think it's a bubble. I think
• 1:02:43
  Speaker 2: it's both of those things. But in any case, if
• 1:02:45
  Speaker 2: they're giving us the opening to restart negotiations of the
• 1:02:49
  Speaker 2: Social Contract, I think we should take them up on that. Yes, indeed,
• 1:02:52
  Speaker 2: all right, guys, that doesn't for us. We've got a
• 1:02:54
  Speaker 2: great Friday show planned for you tomorrow. I think Rocana
• 1:02:58
  Speaker 2: is joining to talk about his fight with Elon and
• 1:03:01
  Speaker 2: reflect on the week's activities with regard to the DSA
• 1:03:03
  Speaker 2: wins he endorsed, I think Claire and Brad Lander, but
• 1:03:06
  Speaker 2: not Daria Liza, which also interested to ask him about
• 1:03:09
  Speaker 2: his endorsement process. So should be a good one. Ryan,
• 1:03:12
  Speaker 2: Thank you, great to have you, great to see you.
• 1:03:15
  Speaker 3: Good to see you too.
• 1:03:16
  Speaker 2: Yeah, safe travels and I will see you guys tomorrow