The Argument for Letting AI Burn It All Down

Suddenly, and not long ago, our dearest tech industry leaders began to suggest caution. Sam Altman said that AI is in a bubble “for sure,” albeit one formed around “a kernel of truth.” Mark Zuckerberg said an AI bubble “is quite possible,” though “if the models keep on growing in capability year over year and demand keeps growing, then maybe there is no collapse, or something.” Even Eric Schmidt is saying to calm down about artificial general intelligence and focus on competing with China.

The question everyone wants an answer to is: How will the bubble pop? Will we wake up and realize that we don’t really want to talk to LLMs anymore? Will someone find a way to build AI tools at one-thousandth the price, letting a thousand ChatGPTs bloom? Are we going to check the news one day and see those photos of stock traders yelling to each other on the floor of the exchange as tech companies’ stock prices blink bright red? My answer is: I have no earthly idea. But I really, really hope that, someday soon, AI becomes … normal.

I love normal technologies. They come with manuals. They change periodically, but you can build craft and professional skills around them. Bubble technologies change constantly, and there is always a threat that they will either destroy society (bad) or make everyone besides you wealthy (worse). There are many ways to forecast when a technology is becoming normal—price-to-earnings ratios and other boring stuff. The metric I use is the C/B ratio: conferences to blogging. If people are steadily attending conferences about a subject, it is not normal yet. If they’re mostly blogging about it, it is. I made this up, but I assure you it’s predictive.

I work with AI all day long, and right now there are so, so many conferences and gatherings and not that many good, boring technical blog posts. The tech industry loves conferences, because our product is so abstract that it’s hard for us to figure out where we sit in the nerd-chimp hierarchy. This is why VC firms are so often sponsoring get-togethers; they allow for pheromonal exchanges and dominance displays, usually enacted with PowerPoint. Invoke the Chatham House Rule if you’re feeling naughty.

People sometimes talk about the golden age of blogging but less about why people blogged: No one had money, and nothing is cheaper than putting words online. When the money flies to money heaven, and the startups become enddowns, the conference budgets are often the first thing to go. Nerds still want to talk their nerd talk, though. That’s when they start posting—it’s the only way to figure out who you are. Eventually, AI’s C/B ratio will start to tip blogward.

Not yet, though. We may have a ways to go. The globalized economy has become, out of expedience and greed, a world-spanning suspension bridge, hung off a few giant anchorages like OpenAI and Nvidia and Google, bolstered by promises of planetary AI transformation—and if one of those anchorages were to falter, just a little, and the promises fail to materialize, maybe the cable would sag and the whole bridge would crumble, and all the AI startups (including mine) would fall into the sea. Constantly anticipating this is just one of the many things that has made 2025 so fun.