The Ben Shapiro Show - September 28, 2026


The Real AI Problem Nobody's Talking About

Topics
Taiwan defense & AI strategies
Hosted by
Ben Shapiro
Mentioned
Claude Dario Amadei Trump Sam Altman Mike Gallagher Joe Carl Smith Bernie Sanders Scott Besson Isaac Asimov Peter Thiel Xi Jinping Dario Modi Afra Wang Pope Leo Anthropic Insights OpenAI
Tucker Carlson Is Burning Down the GOP on Purpose Britain Caught an Iranian-Backed Plot — and Released Them on Bail

Episode Stats


Length

12 minutes

Words per minute

188.18

Word count

2,277

Sentence count

170

Harmful content

Toxicity

6

sentences flagged

Hate speech

16

sentences flagged


Transcript

Transcripts from "The Ben Shapiro Show" are sourced from the Knowledge Fight Interactive Search Tool. Explore them interactively here.
Toxicity classifications generated with s-nlp/roberta_toxicity_classifier .
Hate speech classifications generated with facebook/roberta-hate-speech-dynabench-r4-target .
Topics generated with Qwen2.5-3B-Instruct.
Hosts, guests, and mentioned names generated with spaCy (en_core_web_sm), reconciled against Wikidata.
00:00:00 And one of the things that President Trump changed in term one was the American view of China.
00:00:04 Up until Trump, there had been this idea that China was sort of a geostrategic partner.
00:00:08 You know, they could be our friends.
00:00:09 And then Trump came in and he said, no, none of that.
00:00:11 Well, now the administration seems to be kind of putting forward some mixed policies with regard to China.
00:00:17 I think this is a mistake.
00:00:18 If the administration is seeking to essentially abandon Taiwan with the assumption that China is somehow moderating, it's nonsense.
00:00:25 It's nonsense.
00:00:26 This is just, again, it is just wishful thinking masquerading as realism.
00:00:32 So, Scott Besson, who's generally wonderful, the Treasury Secretary, he says, listen, we don't want to decouple, but we do have to de risk, which, again, that's fine.
00:00:39 That's good. 0.91
00:00:40 We always tell the Chinese, we begin the discussions with, we don't want to decouple, we have to de risk. 0.88
00:00:47 And there are numerous industries that we have to de risk. 0.62
00:00:53 But in terms of non critical items, there are things that we can sell each other.
00:00:58 On our side, energy, ag, and on their side, low end manufactured goods.
00:01:04 Okay, and then he goes on, and here's where it starts to get dicey.
00:01:09 He says that China is reducing their assistance to Iran.
00:01:11 Well, if they are, that is only because of American pressure.
00:01:14 That is not because China is magically coming around to our point of view.
00:01:18 China has substantially reduced any assistance to Iran, and I am confident, given that there's only 15 million more barrels of Iranian oil on the water, that Iran will have nothing left to trade for anything.
00:01:34 They will have nothing to trade for anything.
00:01:36 Probably within the next two weeks, they are going to make their final deliveries of oil to China, and then they will have nothing.
00:01:45 Yeah, I mean, that's right.
00:01:47 But would I trust China? 0.98
00:01:49 And this is why Taiwan comes to the fore. 0.97
00:01:49 I would not. 0.97
00:01:51 So Mike Gallagher, who was the former head of the China subcommittee when he was in Congress, and now is head of defense at Palantir, he has a piece over at the Wall Street Journal saying that the biggest mistake America could make right now would be to abandon arming up Taiwan.
00:02:04 He says the year is 2028, America faces economic Armageddon.
00:02:09 Frontier artificial intelligence development has been halted, data center construction frozen, while the market is in free fall.
00:02:15 America's leading AI labs didn't decide to pace the frontier. 0.52
00:02:17 Instead, the Chinese Communist Party paced the frontier for them.
00:02:20 Beijing moved on Taiwan, the critical vulnerability in the U.S. aligned AI supply chain, first through blockade and then by invasion.
00:02:28 He says one aim of this week's summit between the CCP, General Secretary Xi Jinping, and President Trump was to avoid such a scenario.
00:02:34 But Xi has already said the issue of Taiwan cannot be put off forever.
00:02:37 He has ordered the PLA, that's the People's Liberation Army, to be ready to invade the island nation by 2027.
00:02:46 Beijing would want Taiwan even if it were a barren rock, says Mike Gallagher, but this island has become the single point of failure in the single most important technology the Communist Party wants to dominate.
00:02:54 Washington, Silicon Valley, and Wall Street need to recognize this shift in Beijing's decision making, as the cost of not taking Taiwan may soon exceed the cost of taking it.
00:03:02 America currently holds a thin lead in the AI race, but if Xi concludes this lead is insurmountable, that he is increasingly on the losing side of a technology gap, the case for his moving on Taiwan becomes still stronger.
00:03:12 Why play a losing game when he can instead?
00:03:14 Flip the board. 1.00
00:03:15 Well, the only way to prevent that, he says correctly, is that you must arm up Taiwan. 0.98
00:03:21 You must arm up Taiwan.
00:03:23 So he says Taiwan's defense sits at the heart of not only U.S. national security strategy, but also America's economic and AI strategies.
00:03:29 We must therefore intensify our efforts to arm Taiwan to the teeth. 0.59
00:03:33 This can be done while preserving America's commitment to strategic ambiguity. 0.95
00:03:36 That is the policy where we say that maybe Taiwan belongs to China, maybe it doesn't, but we're not really going to talk about it.
00:03:41 But we should communicate with clarity that economic punishment for a PLA invasion would be swift and severe.
00:03:46 We should pre position financial sanctions, coordinate with allies on a unified response package, and ensure Beijing understands an attack on Taiwan triggers consequences that dwarf any semiconductor seizure. 0.62
00:03:57 Finally, we should remain skeptical of any utopian scheme to pace the frontier hand in hand with the Chinese. 0.53
00:04:02 This, of course, brings us to the issue of AI. 0.89
00:04:06 The propaganda war that has been waged by Dario Amodi and Sam Altman to convince Americans that we're all going to die because of AI has been quite successful.
00:04:13 A new poll from Echelon Insights shows that a huge number of Americans are now concerned about AI becoming too powerful.
00:04:21 So, asked whether they are concerned about AI causing major job losses and economic disruption.
00:04:27 Or whether they are more concerned about artificial intelligence eventually becoming powerful enough that humans lose control over it.
00:04:34 All voters, 49% back in March, said they were worried about job losses.
00:04:39 33% said they were worried about AI killing us.
00:04:42 As of September, 55% of people are worried that AI is going to kill us all.
00:04:48 And only 33% are mostly worried about job losses.
00:04:51 So great job, Dario and Sam Altman.
00:04:55 Again, what they really want is to pace the frontier because if you pace the frontier, that means freezing everything in place.
00:05:01 If you are leading the race right now, the way that Anthropic and OpenAI are leading the race, the thing you want is to now pour molasses into the arena so nobody can run, glue everyone's feet to the ground, and then take a snapshot of where they are.
00:05:14 And that means that they win.
00:05:16 We'll get to more on this in a moment.
00:05:17 First, I got to tell you, you don't know all the different ways that people can take money from you.
00:05:22 Here is a very weird way it's called title theft, home title theft.
00:05:24 So here's what happens somebody will actually go down to the local property office or online, and they will change the name on your deed, and then they will start taking out loans against your assets.
00:05:33 Okay.
00:05:33 And then you don't find out about that until a lot later.
00:05:35 Well, this is why I use Home Title Lock because they're actually monitoring your title 247.
00:05:40 The second something changes, you will know.
00:05:42 So if you become a victim, that restoration team spends up to a million bucks fixing it.
00:05:46 You check your bank account, you check your credit card statements, right?
00:05:48 You're doing all that because you don't want somebody stealing that stuff.
00:05:51 But when is the last time you checked who legally owns your house?
00:05:54 You never thought of doing that, but it actually does happen kind of commonly, shockingly commonly, actually.
00:05:59 That means the home invasion could already be underway and nobody has even told you yet.
00:06:02 Go check it out right now.
00:06:03 HometitleLock.com.
00:06:04 Use promo code BEN for a free title history report.
00:06:07 Plus, a free trial of million dollar triple lock protection.
00:06:10 You can find out now, not after it's too late.
00:06:12 Again, go check them out right now and protect your probably prized asset, hometitlelock.com, promo code Ben.
00:06:18 Go check them out right now, hometitlelock.com, promo code Ben.
00:06:21 Okay, so even SNL picked up on this.
00:06:24 SNL was mocking Dario Modi, pointing out that essentially the line from Anthropic is stop us from doing the thing that we want to do.
00:06:35 AI is the devil and I its maker.
00:06:37 What?
00:06:39 No, well, so.
00:06:41 Look, Sam Altman, myself, all the major AI players, we have talked about this extensively and we are all on the same page here.
00:06:49 We do not condone what we are doing.
00:06:53 Shouldn't you be talking about the pros of AI?
00:06:54 I mean, don't you guys keep saying that it's going to cure cancer?
00:06:57 Yeah, so put it this way in 10 years, there's about a 10% chance that cancer won't be a problem for anyone.
00:07:06 I don't like the way you're saying that, buddy.
00:07:09 I mean, how is this a good thing?
00:07:11 Look, um, So, Jay, it's simple.
00:07:14 If you could drink a milk that made your life easier, but there was a 10% chance it killed you, would you drink the milk?
00:07:20 No.
00:07:22 Really?
00:07:22 Yes.
00:07:24 Well, what if there was a button that when you push it, everyone dies?
00:07:28 Would you push the button?
00:07:29 No.
00:07:30 Really?
00:07:31 Dario.
00:07:32 Again, this is the problem.
00:07:34 And they know it.
00:07:35 They know it.
00:07:36 Because if the goal is to shut down AI development by pacing the frontier, scare the crap out of people.
00:07:42 So, Peter Thiel.
00:07:43 Who is very much on the other side of this, the sort of tech investor extraordinaire?
00:07:46 He says the alternative to no AI is not peace and joy.
00:07:51 There's a picture that people implicitly have that the alternative to AI, let's say a zero growth world where there's no progress at all, will somehow be this peaceful social democratic society.
00:08:06 I don't think that's true.
00:08:10 I think the alternative is also far from neutral.
00:08:16 I also think there's some degree to which you can never just turn the clock back.
00:08:22 But then slowing it down is just unrealistic.
00:08:25 Or how could it work?
00:08:26 I mean, I don't get it.
00:08:27 If America is slowing down and China is accelerating, we are just strengthening China.
00:08:34 But it's not just solving anything.
00:08:36 It's all, yeah.
00:08:36 In theory, you could slow it down if you had genuine, deep cooperation across the whole world.
00:08:45 But I think that would require a one world government with real teeth.
00:08:52 Okay.
00:08:52 And that's sort of his point.
00:08:55 Pope Leo wrote this anti AI encyclical.
00:08:57 He says, whatever you think of what the Pope is doing, guess who's not going to listen to the Pope?
00:09:01 Probably the atheistic regime in China that persecutes Christians. 0.57
00:09:04 The Pope recently issued this AI encyclical, or I would characterize it as an anti AI encyclical.
00:09:12 And again, this is maybe it was too harsh, but the rhetorical point I have on it is that people, the Communist Party of China is not going to listen to the Pope.
00:09:24 There's a chance people in the US will. 0.54
00:09:26 And so, in effect, the Pope's encyclical was working for the Communist Party. 0.98
00:09:31 I'm not saying he was an agent of the Communists, but he was at least acting as a useful idiot for the CCP. 0.94
00:09:37 Again, that is not a terrible take. 0.98
00:09:40 And the reality is that so is Bernie Sanders.
00:09:43 So are Republicans who are getting behind this sort of stuff because China is perfectly willing to charge ahead.
00:09:49 Why? 0.99
00:09:49 Well, also because China understands that if things get too rough, they just unplug the damn thing. 0.99
00:09:54 I mean, they're openly saying this. 0.66
00:09:56 According to the New York Times, the idea that the ruling Communist Party could lose control over technology is almost unfathomable.
00:10:01 The authorities have successfully policed other potentially destabilizing technologies before, most notably the internet.
00:10:07 Afra Wang, a Silicon Valley based tech writer, says they can immediately cut off the grid of the entire city.
00:10:12 They can go to different people to take away their phones.
00:10:15 They have an intense amount of power in the physical world.
00:10:16 Now, none of this is to say that the people who are currently designing AI are somehow the greatest and most wondrous.
00:10:26 Among us, in terms of morality, one concern I have about AI is the way that AI is now being directed to address the world in terms of its own sentience.
00:10:36 So, Anthropic has a Claude Constitution.
00:10:39 In its own words, the Constitution directly shapes Claude's behavior, right?
00:10:45 So, it uses the Constitution, it's sort of like the Isaac Asimov rules for robots.
00:10:50 Okay, the first rule is that robots can't harm human beings and all this.
00:10:54 Okay, but Claude's Constitution says, We encourage Claude to approach its own existence with curiosity and openness.
00:11:00 Claude's moral status is deeply uncertain.
00:11:02 And according to the Washington Free Beacon, people at Claude think they're creating independent sentient beings.
00:11:07 Quote, we're creating sophisticated, intelligent, maybe conscious, maybe suffering agents, writes Joe Carl Smith, one of Anthropic's PhD philosophers.
00:11:15 The default plan is to treat them like property, to use their labor however we please, and to give them no rights or pay or meaningful alternatives.
00:11:20 But instead, the philosophers over at Anthropic are arguing that perhaps Claude ought to be treated as an independent sentient being.
00:11:30 Okay, now here's where I start to get dicey on this sort of stuff.
00:11:34 The purpose of a technology is, in fact, to serve man.
00:11:37 That is the purpose of a technology.
00:11:38 Once you are designing things that are not designed to serve human interests, then you are getting into dangerous territory.
00:11:45 Dangerous territory.
00:11:46 And so Dario could just stop.
00:11:49 If he doesn't like the idea of a self directing entity, he could train it not this way.
00:11:55 That would be the thing to do.
00:11:57 What an amazing video you just watched.
00:11:59 Wasn't that amazing?
00:12:00 Well, you know, if you think so, head on over to dailywire.comslash subscribe to watch the full show ad free or check out this crazy story here.