**Andrew Keen** (0:00)
Hello, my name is Andrew Keen. Welcome to Keen On America, the daily interview show about the United States.
Hello, everybody, and welcome. It's Saturday, June the 20th, 2026 And it has been one of those weeks where the machines quietly got more capable and the rest of us got a little more nervous about our own job descriptions. Space. Hello, everybody. It isn't actually June 20th. It's June the 27th. That was the show that Keith ran last week, featuring a fake Andrew Keen. So Keith, that was a very realistic version of my voice. I've become redundant. You don't need me anymore, do you?
**Keith Teare** (1:06)
Well, there's only one you, Andrew.
**Andrew Keen** (1:08)
And clearly not only one me. You've invented another one that's just as credible and probably more interesting.
**Keith Teare** (1:15)
So everyone should know that Andrew was away last week and I wasn't going to do a video. And it suddenly occurred to me, I've got a PC here that has an Nvidia card.
I didn't use any third party service, actually. I just said to it, look, here's the last few That Was The Weeks. Learn Andrew's voice.
And here's an image of him from YouTube. Animate Andrew, write a script based on this week's newsletter, and animate Andrew and me discussing it. And it took about half an hour and maybe three attempts because he wasn't very good the first time. But after about an hour's work, that came out. It was actually pretty bad compared to our normal actual live shows, but it was really good.
**Andrew Keen** (2:03)
Well, when I heard it the first time, it sounded like myself. In fact, I second guessed myself. I thought maybe I really was there on June 20th. Anyway, I wasn't. Last week, you did a show on your own. This week, it's Saturday, June 27th. The big news seemed to have happened yesterday, Keith. The Washington Post reports that the US government will decide who gets to use the latest AI technology.
Commerce Department apparently told both OpenAI and Anthropic that they needed their permission about who was going to get to use their latest model. As your editorial this week announces, Payback's a Bitch, Baby. Of course, you blame everything on Dario, do you, Keith? Does he finally get what he deserves?
**Keith Teare** (2:50)
It's definitely not only Dario.
We've talked about the crisis of leadership in AI across really the whole of AI. The quality of the leadership is questionable at best. There's been a year now, more than a year of all of the AI leaders at different times crying wolf about how dangerous it's going to become because they wanted to be wrapped in government clothing to prevent third parties being able to easily compete with them. In a way, the government provides a moat of protection against anyone else, if you have to go through government in order to do something. That's what they, especially Dario, but not only Dario has wanted, it's now happened. And now they're, you know, I'm pretty sure they realize that what they've asked for isn't that really a good thing. They've given the government authority. And by the way, it is voluntary right now. There's no law, so they could ignore the government.
Law has to be written or an executive order has to be written. And I do believe that is in process, but it hasn't happened yet.
**Andrew Keen** (4:05)
Yeah, and your editorial says, the government is not merely getting involved in the conversation about AI. That's been true for years. No, the Frontier AI companies invited the government into the room, and now the government is beginning to behave as if it owns the door, the guest list, the schedule and the product roadmap. In other words, the AI Frontier companies like Anthropic and OpenAI wanted their cake and eat it.
And they didn't get to do either. They've had the cake taken away. I wonder, Keith, if we step back a little bit. We've talked about this theme before, whether we might think a little bit more structurally and see the convergence of the Chinese and American model. And yeah, you're probably right in the sense that Dario and Sam Altman aren't exactly the most mature of leaders, although they're tech people. They're not political people. But this was almost inevitable, particularly in the context of Trumpism.
**Keith Teare** (5:07)
I don't think it was inevitable, actually. I think it's been set up.
It became more and more likely, and now it's happened. I think it's pretty inevitable that it's the start of a pretty onerous framework that they're going to have to live within. You're right about the Chinese models. GLM 5.2 was released this week, which is as good as Claude, and is open source, open way. There are some people who think it's a very big model and very capable, but there are some people...
22 more minutes of transcript below
Try it now — copy, paste, done:
curl -H "x-api-key: pt_demo" \
https://spoken.md/transcripts/1000651996090
Works with Claude, ChatGPT, Cursor, and any agent that makes HTTP calls.
From $0.10 per transcript. No subscription. Credits never expire.
Using your own key:
curl -H "x-api-key: YOUR_KEY" \
https://spoken.md/transcripts/1000774487577