The AI optionality issue

There has been quite a lot of talk about the AI bubble. The causes are complex, and yet, quite simple also – Investors have put too much money into unsustainable beliefs around AI. I’m not economics expert, so I can’t comment on that side. But there is a familiar problem with the investment in AI (both financial and emotional), which we’ve seen with many previous technologies.

The issue is that for this kind of investment to be realised, AI needs to become ubiquitous, to become essential to the general population. This was the problem with the Metaverse. While it could be useful for specific applications, particularly in education, to warrant the investment required it needed to be a universal solution for everyone – where you did your shopping, socialising and learning. And it just wasn’t that (an aside – we don’t see those adds for a Metaverse utopia anymore do we?)

AI is undoubtedly more useful than the Metaverse for more people. But it’s not _that_ essential. Let’s play a thought experiment. Imagine AI (by which I mean the LLM services such as ChatGPT, Claude, Gemini, etc, not specialised applications in medicine or whatever) was turned off tomorrow. How much would that impact your practice? There are undoubtedly some people who have come to rely on (all those men with AI girlfriends would be inconsolable), but for most of us I suspect it would be a tsk or annoyance, or even just a gallic shrug, but it wouldn’t be a fundamental disruption to practice. Compare that to the following technologies suddenly disappearing:

• Mobile phones
• Streaming services for TV and music
• Podcasts/Audiobooks
• Online services (banking, Governmental)
• Social Media
• Online retail

Now, for any one of you, not all of these will be important, but I would suggest that quite a few of them would be more inconvenient to lose than AI. An obvious objection to make to my list is that these technologies have had a lot longer to establish themselves in our lives. That is true of course, but it is also the case that they were instantly useful. And the problem is that the type of investment that AI is currently scooping into its maw can’t wait for 20 years to see a return. It probably can’t wait 20 months.

AI can be fun, and is useful in places, but for far too great a proportion of the general population it is entirely optional. This is why the drive currently is to make it mandatory, to embed it everywhere, so by the time the collapse looms, it is pushed into the interstices of everyday life to make it inevitable (Doctorow’s phrase of “AI is the asbestos we are shoveling into the walls of our society and our descendants will be digging it out for generations” is powerful). For the investors it is a race against time – can they make it essential by stuffing it into things like MS Office before the bubble bursts. Without that pressure the optionality of AI will be the undoing of the investors. Too many of us can take it or leave it. Like beige carpet.

This work is licensed under a Creative Commons Attribution 4.0 International