AI Bubble Paradox

Most Vikings were women, children, or elderly. Even among the able-bodied men, most were farmers, or fishermen, or something non-axe wielding. But let's be honest, when you hear "Viking" you think blood eagles and murderous warriors hellbent on dying in battle to reach Valhalla.

Take a quick logic breath. The Vikings didn't win so many battles by dying in them, they won by making the other side do the dying. Quite how many actually jumped off their longboats hoping to be impaled on a Saxon spear will be forever unknown, but it probably wasn't many. Most, I imagine, were focused on living long enough to sell the slaves and loot, and buy themselves lives of luxury. Or at least what counted as luxury before central heating, antibiotics, and toilet paper.

Still, death wishes and scantily clad Valkyries make excellent poetry for the sagas. Not to mention a charming contradiction in terms — the Vikings survived so long by praying for death. It's almost the reverse of those industries that would fail if they ever truly succeeded. Smoking cessation products, for instance. Condoms as well, to a large extent because if everyone used them, we'd eradicate communicable groin problems (and trigger a baby boom). I'm going to segue into the AI hype bubble.

And a bubble has been declared. Sundar Pichai, Jensen Hwang, and Sam Altman have all made public statements to that effect. So, we can safely assume that by now, everybody who should know we're in a bubble — i.e. Valley bigwigs and financiers — does know. It's still growing. Every big tech company is now an AI company with trillions of extra inbound investment. Construction companies are harvesting the windfall of giant datacentres being built. Energy conglomerates are rubbing their oil and blood soaked hands together at the prospect of powering it all, presumably as their executives club baby seals and drown polar bears for a bit of weekend fun. And there is a mind-boggling network of circular investments.

Everybody in tech, or just adjacent to it, is investing in everybody else. Then, they're taking the money they've received — in the form of increased valuations — to make more investments in each other and grow those valuations even higher. Repeat, ad infinitum. We are in a bubble. Worse, we're dependent on it. Almost every cent of growth in the US economy has come from the valuations of AI companies. China is banking on AI to lift it from its productivity woes, and those woes don't even count as faint scratches compared to the slumps across Europe. All of us are making one giant bet that AI will fix our economies.

Take another quick logic breath. The AI we've seen over the last few years is mesmerizing. Before it came onto the scene, I turned away a massive international client who wanted to mine textual clinical data for operationally useful information. That was impossible, then. Now, we can improve patients' lives by doing it. AI is a good thing. But the way I'm viewing AI isn't a hugely profitable vantage point. For that, players across the economy have to see it bumping up their bottom lines in one of two ways: 1) creative uses of the technology that open new means of production, or 2) cost savings. The first is hard, the second is laying off human workers. And practically, everyone with the power to choose is planning on option 2. We're collectively betting on the economy being rescued by technology-induced redundancies.

I can't think of a single historical precedent for this lunacy. Unless, perhaps, the Vikings really did survive by dying.