Web dev at the end of the world, from Hveragerði, Iceland

Bubbles are exhausting

I’ve been spending the last couple of weeks getting started on the plan I formed after my epic eight part strategy retrospective and doing odd jobs for a couple of clients that have resurfaced.

It’s fun. Especially the part where I try to figure out the simplest possible way of executing on the plan I formed using tools that have appeared (or matured) in the time that has passed since the last time I rethought my web presence.

That was, I think, seven or eight years ago. People have made so many interesting tools in the meantime.

What’s exhausting is when I surface, and everywhere I see is just endless bubble talk.

  • Fact-free assertions of how amazing generative models are.
  • Selective readings of history used to dismiss valid concerns.
  • Science-fiction claims on what generative models can do.
  • Theatrical hand-wringing based on those science-fiction claims and not the valid concerns.
  • Self-avowed sceptics who turn around and start inflating the bubble themselves.

Most people involved are relying on second-hand assertions of vendor studies tailored for marketing and sales. Their echo chamber repeats the claims until they petrify into ‘truths’. To paraphrase Nietzsche, the ideas of the AI Bubble have become a mobile army of metaphors, metonyms, and anthropomorphisms. They’ve constructed a reality and now believe that their embellishments are truth.

The only way to counter endemic myth-making is scepticism. Most of the claims fall apart as soon as you try to validate them and the rest dissolve because they are pure fantasy – speculation that has hardened into truth through repetition and hearsay.

This constant effort is exhausting and after a while you end up having to filter out the ‘usual suspects’ – the people who keep repeating falsehoods and hyperbolic claims.

That work, too, is exhausting.

All of this is just how bubbles work. The crypto-coin bubble, though smaller, was exactly like this.

The financial bubble leading to the 2008 crash? Same.

Web 2.0? Yup.

Dot-com? Yup yup yup.

Some of these were genuine innovations. Some of these were just fraud.

Even in the unlikely event that “AI” is the revolution supporters claim, that only means that the bubble will be even largereven more destructive. That it’s mostly fraud and shitty automation is the optimistic take. The alternative is that we’re going to get the largest, most energy-intensive financial bubble in history right when we need a concerted global effort to prevent the global climate crisis from tearing the very fabric of modern society apart.

I prefer to be an optimist and think that the AI industry is continuing its pattern of over-promising and outright fraud., and that Microsoft and Google are continuing their pattern of extracting, not creating, value from captured markets. What I’ve seen so far and the research I’ve done strongly suggests that this is the case.

The alternative is also just too ghastly consider.

You can also find me on Mastodon and Twitter