Gmar chatima tova! There’s a story I like (probably apocryphal) that relates to Yom Kippur: Rabbi Eliezer famously told his students to repent one day before they die. When they reasonably pointed out that no one knows when that will be, he told them to repent today. A clarifying reminder as we go into a meaningful and easy fast

The political implications of the coming AI IPOs are somehow still underappreciated Historically, each wave of wealth (railroads and steel, oil, finance, early tech, etc) was accompanied by a new class of people for whom politics was a way to compete for status and finance their convictions AI is likely to follow the same path. If the major labs go public over the next year or so, their founders and early employees will gain access to personal liquidity extraordinary in both its breadth (the number of employees) and depth (the absolute size of the wealth). Beyond simply financing candidates, they will be sustaining institutions and underwriting political projects for decades There need not be a coherent set of “AI politics” for this to matter (in fact, I doubt there is one!). These people can disagree profoundly and politics will still proactively warp itself in their direction. Whether you welcome this is a separate question. But the smart players in DC are building these relationships already

One funny part about how DC works is that government often reaches for off the shelf policy in moments of crisis So it stands to reason that Washington will likely discover an enormous appetite for biodefense immediately after a potential biological attack But today, the only available policy solution is restricting AI, which is far easier than deploying biodefenses post hoc We should probably have more options — and certainly more effective ones — ready before then

Every cycle someone gets famous for "blowing the whistle" on tech (Meta then, OAI/Ant now) It's not an op—there's just an endless appetite from the liberal New York Times-class for anti-tech testimony Unfortunately the sincere idealism of lab employees plays right into it

The biggest debate of the next decade will be between people who look at this and love it — and people who see this and desperately want to stop it If you thought the expertise/"misinformation" debate was bad during COVID buckle up