Spike Lee was paraded through Fort Greene in Brooklyn like he was the pope.

Our mayor is Muslim Our bagels are Jewish I feel so damn alive CUZ THE KNICKS JUST WON IN FIVE

Nah, he's alright. Let that boy do his thing.

Wow, I can't believe he drew this. Incredible. Genius. Ahead of his time.

1. The economics do not work. Cutting off the world isn't survivable — for Anthropic or for US tech. Start with the macro: however you measure, the US is ~26% of world GDP (nominal) or ~15% (PPP, and falling). Most demand sits abroad. statisticstimes.com/economy/…

My Knicks have an opportunity to win The Finals on my birthday tomorrow. Life-changing joy on the line. The mood carries.

There's lots of fear about AI and price discrimination. "They're using data in all sorts of complicated ways we can't even comprehend!" Fair. So instead of assuming a specific form, ask, what can happen across all possible types of data? Are consumers doomed? I'll argue no.

The revised AICOA lowers the harm threshold, raises the bar for security and privacy defenses, and speeds up litigation. @geoffmanne argues it doesn't fix the bill's core problem: labeling a company "systemically important" doesn't establish consumer harm. Full statement ⬇️

AICOA is back... The bill restricts self-preferencing, like when platforms favor their own services. Google Maps in Google Search. Amazon Basics on Amazon. Sure, competitors hate this. But competitor harm is not the same as consumers being harmed. This is antitrust 101 stuff

Everyone always talking about “talent density” in Silicon Valley when we really should be talking about how 80% of pretzels in America come from a small region of Pennsylvania

#Competition #DigitalMarkets Apparently, some US politicians don't want Europe to have a monopoly on bad regulation... so "AICOA" is back from the grave. Sens. Grassley and Klobuchar have introduced a new American Innovation and Choice Online Act (link in the first comments). It is narrower than the original version in whom it targets—but, in important respects, more aggressive in how it regulates them. As @geoffmanne has said, “AICOA has always been a solution in search of a problem”. Why regulate digital markets? Digital markets have always been very competitive. So-called "big tech" firms have expanded output and invested heavily in R D. Consumers have benefited from constant innovation. Even firms with clear leadership in some markets face competition from disruptors or other big tech firms entering adjacent markets. Tech firms explain, at least in part, the productivity gap between the US and Europe... so why imitate European regulation? Why *now*? Regulation makes even less sense in the midst of the generative-AI boom. GenAI is already reshaping—and could fundamentally disrupt—search, social media, e-commerce, and other digital markets. The revised AICOA targets a small group of “systemically important platforms” based largely on size and reach, not proof of market power or consumer harm. It would treat practices such as self-preferencing, integration, defaults, ranking, data use, and interoperability restrictions as presumptively suspect, even though these practices can improve convenience, security, and product quality. Its low “material harm” threshold and demanding defenses would make ordinary design and engineering decisions far riskier. The likely result is more litigation, slower product improvements, and less integration—not necessarily more competition. Large penalties and expedited enforcement would further encourage overcompliance. Governments should address exclusionary conduct when there is evidence of actual competitive harm (and we have Antitrust Law for that), rather than adopting a DMA-style shortcut that risks protecting competitors at the expense of consumers and innovation. While I enjoy zombie movies, zombie bills are a waste of valuable resources that we should not entertain.

A lot of people waiting for the official @GusHurwitz statement on AICOA reintroduction. Wait no more...

If you were wondering what the "pause" was all about, Ben Thompson @stratechery has an interesting theory: "Late last week the Anthropic Institute released a new safety report warning about the danger of recursive self-improvement... I don’t think the timing is a coincidence. This is a company and leadership that has been honing safety-and-scaremongering-as-marketing-tactic ever since Amodei led the charge to close source OpenAI models because GPT-2 was too dangerous; it’s always fun to see the evolution of tactics, capabilities, and goals, and in this case publishing a widely-discussed report the week before you cite it to silently degrade your offering for potential competitors is impressive."

The speech is worth a read, but a central premise requires further explanation: "We should prevent AI from taking over jobs that only humans ought to do..." The word "ought" is doing all of the work here, work that should be explored further so the notion can be universalized and discussed for what it likely is - a rent seeker's dream.

“Cue the quote tweets.”

Yesterday, Commission spokesperson Thomas Regnier gave Brussels' official line on Apple withholding Siri AI from EU iPhones. It's a remarkable statement—and nearly every clause of it gets the economics backwards. Let's go through it, clause by clause. 🧵

When regulation makes a product legally or commercially untenable, the binding constraint is the regulation, not the technology. @AuerDirk explains why the DMA left Apple choosing between privacy risks, penalties, or withholding Siri AI from EU users. Read below ⬇️

Or it’s simply poor regulatory design. You can’t legislate every outcome into being. Force companies to bankroll rivals and degrade their own products, and the rational and legitimate response may be to not ship in that market at all.