The Argument for Letting AI Burn It All Down

Suddenly, and no Long ago, our dearest tech industry leaders began to suggest caution. Sam Altman said that AI is “of course” in a bubble, even if it is formed around a “kernel of truth”. Mark Zuckerberg said an AI bubble “is entirely possible,” but “if the models continue to grow in capacity year over year and demand continues to grow, then maybe there won’t be a collapse, or anything like that.” Even Eric Schmidt recommends calming down about artificial general intelligence and focusing on competition with China.

The question everyone wants to answer is: how will the bubble burst? Will we wake up and realize that we don’t really want to talk to LLMs anymore? Will anyone find a way to create AI tools for a thousandth of the price, allowing a thousand ChatGPTs to flourish? Will we ever look at the news and see those photos of stock traders shouting at each other on the trading floor while the stock prices of technology companies flash bright red? My answer is: I have no earthly idea. But I really, really hope that one day soon, AI will become… normal.

I like normal technologies. They come with manuals. They change periodically, but you can develop craft and professional skills around them. Bubble technologies are constantly evolving and there is always a threat that they will destroy society (bad) or make everyone rich (worse). There are many ways to predict when a technology will become normal: price-to-earnings ratios and other boring things. The metric I use is the P/B ratio: conferences/blogging. If people regularly attend conferences on a topic, it’s still not normal. If they blog primarily about it, that’s it. I made that up, but I assure you it’s predictive.

I work with AI all day, and right now there are so many conferences and gatherings and not a lot of good, boring technical blog posts. The technology industry love conferences, because our product is so abstract that it’s difficult for us to determine where we fit in the nerd-chimpanzee hierarchy. This is why venture capital firms so often sponsor meetings; they allow for pheromone exchanges and displays of dominance, usually staged 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 money flies to money heaven and startups become management companies, conference budgets are often the first thing to go. However, nerds still want to talk about their nerdiness. That’s when they start posting: it’s the only way to find out who you are. Eventually, the P/E ratio of AI will start to lean towards the blog.

But not yet. We may have a way to go. The globalized economy has become, through opportunism and greed, a global suspension bridge, suspended from a few giant anchors like OpenAI, Nvidia and Google, supported by promises of planetary AI transformation – and if one of these anchors were to weaken, just a little, and the promises did not materialize, perhaps the cable would sag and the whole bridge would collapse, and all the startups of AI (including mine) would fall into the sea. Constantly anticipating this is just one of the many things that made 2025 so much fun.