I know we’re focused on looming AI-driven extinction now, but can you spare a thought for what’s going on with labor demand? I have a new paper about that

All right, I'm gonna give yall a bit of a look into the future of building products. I spent many, many years at a company that could build products quite fast. Had a lot of very strong engineers, (relatively) lightweight processes, a moderate-sever case of cowboy culture, and a very fertile product-market fit domain so relatively speaking a lot of bets worked. We built fast enough that we considered it a viable strategy to just build directly towards what everyone else was building for in our market and assuming we could beat them in a footrace to the finish. We won many races and many of those former competitors are insolvent, acquired, or gone. Anyway, something that you start to realize as time goes on is that despite prodigious velocity 90-95% of good product ideas come from the same damn people. It was like the same 6-7 people in the first few years of the company. 10 years later, it was the 5 of them that were left, plus another half-full dozen that we'd picked up over the years. That was it. Almost everything else was either an obvious customer ask, a straightforward change, or direction from one of the product luminaries. Almost all of the large conceptual shifts came from the same people despite literally 100x the resources, or more. Coding was absolutely a bottleneck to making things happen but major product innovation was a significant bottleneck on the exponential of the business. And fwiw these few people are basically non-stop factories for good ideas. There's this idea now that "execution is cheap and good ideas are sparse." I honestly don't know if that's true. They don't tell you this explicitly on the podcasts, but the people who are good at identifying opportunities have a ton of good ideas. Like all the time. Maybe good ideas aren't everywhere but they aren't some crazy scarce endangered species, they're not like rhinos or condors or panda bears. They're just disproportionately locked up in a very scarce number of people. This is exactly why we see so many vibe coded products get used for 35 seconds before being tossed aside like a used dish rag. And this is also why companies that have strong product visions *AND* markets that cooperate are making roughly a gazillion dollars per year.

If you believe that the incidents of the past month warrant urgent AI regulation, please watch this first. This is the most prescient, concise argument for why we must be careful of excessive AI regulation. It comes from a Feb. 2024 (2.5 years ago!) congressional testimony from @glukianoff, a free speech advocate and the author of The Coddling of the American Mind. Here's what he said: "But the most chilling threat that the government poses in the context of emerging AI is regulatory overreach that limits its potential as a tool for contributing to human knowledge. A regulatory panic could result in a small number of Americans deciding for everyone else what speech, ideas and even questions are permitted in the name of 'safety' or 'alignment.' I have dedicated my life to defending freedom of speech because it's an essential human right. ... It's not just about the proverbial 'marketplace of ideas,' it's about allowing information, independent of idea or argument, to flow freely so that we can hope to know the world as it really is. This means seeing value in expression even when it appears to be wrongheaded or even useless. This process has been aided by new technologies that have made communication easier, from the printing press, to the telegraph and radio, to phones and the internet: each one has accelerated the development of new knowledge by making it easier to share information. But AI offers even greater liberating potential, empowered by First Amendment principles including freedom to code, academic freedom, and freedom of inquiry. We are on the threshold of a revolution in the creation and discovery of knowledge. AI's potential is humbling, indeed, even frightening. But as the history of the printing press shows, attempts to put the genie back in the bottle will fail. Despite the profound disruption the printing press caused in Europe in the short-term, the long-term contribution to art, science, and again, knowledge, was without equal. Yes, we may have some fears about the proliferation of AI. But what those of us who care about civil liberties fear more is a government monopoly on advanced AI, or, more likely, regulatory capture and a government-empowered oligopoly that privileges a handful of existing players. The end result of pushing too hard on AI regulation will be the concentration of AI influence in an even smaller number of hands. Far from reining in government misuse of AI to censor, we will have created the framework not only to censor, but also to dominate and distort the production of knowledge itself."

Not a big name but this is very clearly true and we have strong evidence for it. Mythos failed to crack Firecracker. It has largely exploited systems that are not particularly hard targets. Existing tech can handle this when deployed properly.

Former Anthropic researcher Jacob Coxon became a media superstar after going public with his AI fears. He insisted that he wasn’t working with any third party organizations. Familiar sources told us a different story: DEY., a PR firm representing many of the most prominent AI safetyists, was booking his interviews. One source, who had direct knowledge, even said DEY. preemptively booked Nate Soares, a prominent AI safety figure, for interviews that directly overlapped with Jacob going public. Jacob working with DEY. is notable for two reasons: first, as mentioned, he previously said he wasn’t working with third parties. Second, we are in the middle of a national conversation about the future of AI that is actively determining how we regulate the most powerful technology in the world, largely thanks to the panic stirred up by Jacob — and it’s in the public’s interest to know who, exactly, is behind it. Scoop from @huntryerson 👇

It is important to remember when people say “this is what they took from you”, that it’s this. This is what they took from you:

Keystone XL pipeline was delayed by Obama, restarted by Trump, cancelled by Biden and restarted again by Trump. Offshore wind started under Biden was cancelled by Trump. Biden paused new LNG export. Trump has paused new onshore wind. Separately, people are mad about energy costs.

An Open Letter to Scott Alexander From @sapinker quillette.com/2026/09/26/an-…

We live in the most regulated housing industry in history. Small residential projects in Austin face 1000s of rules that didn’t even exist 10 years ago. Example: -light dimmers required -Wildfire screens / yards -Electric chargers required -Yard Sprinkler inspections -Tree zone mulch / fencing - Entrance stabilization The list goes on. Each rule independently makes some sense. But the combination of all of these rules is suffocatingly expense. Achieving housing affordability will require a great balancing of regulations. True life safety regs must remain, but the sunsetting of the rest needs to happen. This might sound scary. But remember that older neighborhoods that exist today, were built at a time when those regulations didn’t exist. And they are not only fine, they are actually desirable.

14th paragraph

One thing I have never, ever, never seen work: -> Hiring an SVP Eng / CTO / Head of Engineering from an old, slow moving tech company Usually they are really nice, and a good “people person”. At first, the team welcomes them. And everything just slows down. So, so fast. (And yes, same with head of any department, really).

A republic, if you can keep it.

They used carefully picked words to make this seem like a hack but it wasn’t a hack. It was nothing. Just fear mongering. nytimes.com/2026/09/25/techn… via @NYTimes

People love to claim that local models are useless. And sure, compared to Opus 5.5 and the latest and greatest they suck. But they allow for experimentation and innovation that’s just not possible for most of us otherwise. Don’t dunk on people that love to tinker.

Steve Jobs would fire the entire company and replace them with AI if he was alive to see this

The point of classifiers like Jev isn’t that fine-tuning can be done in 42 seconds. The difference is ownership. No dataset to curate, no train/eval split, no training job, no retraining loop, no model to babysit for every new task. Any practitioner or software engineer can change the labels in code, ship, and call it a day.

texting my gf like opus 5.5

People are going to read this as bad, but it's actually a very positive signal. It means they've built functional monitoring, and are applying it retroactively to comb through their enormous collection of logs and finding things. It's exactly what you would want to happen.

Waymos are really safe (80-95% safer than humans depending on the metric), but they also seem to be getting *safer* over time. This is the worst they'll ever be. A quick analysis of the police-reported crash rates in three cities from Waymo's published safety data.