From boom to burst, the AI bubble is only heading in one direction | Artificial

“Are we really in an AI bubble,” asked a reader of last month’s column about the apparently unstoppable rise of Nvidia, “and how would we know?” Good question, so I asked an AI about it and was pointed to Investopedia, which is written by humans who know about this stuff. It told me that a bubble goes through five stages – rather as Elisabeth Kübler-Ross said people do with grief. For investment bubbles, the five stages are displacement, boom, euphoria, profit-taking and panic. So let’s see how this maps on to our experience so far with AI.

First, displacement. That’s easy: it was ChatGPT wot dunnit. When it appeared on 30 November 2022, the world went, well, apeshit. So, everybody realised, this was what all the muttering surrounding AI was about! And people were bewitched by the discovery that you could converse with a machine and it would talk (well, write) back to you in coherent sentences. It was like the moment in the spring of 1993 when people saw Mosaic, the first proper web browser, and suddenly the penny dropped: so this was what that “internet” thingy was for. And then Netscape had its initial public offering in August 1995, when the stock went stratospheric and the first internet bubble started to inflate.

Second stage: boom. The launch of ChatGPT revealed that all the big tech companies had actually been playing with this AI stuff for years but had been too scared to tell the world because of the technology’s intrinsic flakiness. Once OpenAI, ChatGPT’s maker, had let the cat out of the bag, though, fomo (fear of missing out) ruled. And there was alarm because the other companies realised that Microsoft had stolen a march on them by quietly investing in OpenAI and in so doing had gained privileged access to the powerful GPT-4 large multimodal model. Satya Nadella, the Microsoft boss, incautiously let slip that his intention had been to make Google “dance”. If that indeed was his plan, it worked: Google, which had thought of itself as a leader in machine learning, released its Bard chatbot before it was ready and retreated amid hoots of derision.

But the excitement also triggered stirrings in the tech undergrowth and suddenly we saw a mushrooming of startups founded by entrepreneurs who saw the tech companies’ big “foundation” models as platforms on which new things could be built – much as entrepreneurs once saw the web as such a foundational base. These seedlings were funded by venture capitalists in time-honoured fashion, but some of them received large investments from both tech companies and corporations such as Nvidia that were making the hardware on which an AI future can supposedly be built.

The third stage of the cycle – euphoria – is the one we’re now in. Caution has been thrown to the winds and ostensibly rational companies are gambling colossal amounts of money on AI. Sam Altman, the boss of OpenAI, started talking about raising $7tn from Middle Eastern petrostates for a big push that would create AGI (artificial general intelligence). He’s also hedging his bets by teaming up with Microsoft to spend $100bn on building the Stargate supercomputer. All this seems to be based on an article of faith; namely, that all that is needed to create superintelligent machines is (a) infinitely more data and (b) infinitely more computing power. And the strange thing is that at the moment the world seems to be taking these fantasies at face value.

Which brings us to stage four of the cycle: profit-taking. This is where canny operators spot that the process is becoming unhinged and start to get out before the bubble bursts. Since nobody is making real money yet from AI except those that build the hardware, there are precious few profits to take, save perhaps for those who own shares in Nvidia or Apple, Amazon, Meta, Microsoft and Alphabet (nee Google). This generative AI turns out to be great at spending money, but not at producing returns on investment.

Stage five – panic – lies ahead. At some stage a bubble gets punctured and a rapid downward curve begins as people frantically try to get out while they can. It’s not clear what will trigger this process in the AI case. It could be that governments eventually tire of having uncontrollable corporate behemoths running loose with investors’ money. Or that shareholders come to the same conclusion. Or that it finally dawns on us that AI technology is an environmental disaster in the making; the planet cannot be paved with datacentres.

But it will burst: nothing grows exponentially for ever. So, going back to that original question: are we caught in an AI bubble? Is the pope a Catholic?

What I’ve been reading

Creation theory
The State of the Culture, 2024 is an intriguing blogpost by Ted Gioia.

Pandemic awareness
The Cambridge Econometrics consultancy’s report The Economic Burden of Long Covid in the UK makes for sobering reading.

Mortal thoughts
Doc Searls, one of the elders of the web, has posted the striking Death Is a Feature on his eponymous blog.