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

I’m available as a consultant. I've written a book on software development and another on the risks of generative AI.

When you promise an AI revolution, eventually you will have to deliver


I’m looking forward to an explosion in features, functionality, and quality apps in a few months: a wealth of new apps, new features growing over software like mould over old bread, every company churning out wins by the truckload.

Because apparently AI is a coding productivity multiplier, revolution in how we make software, and absolutely, definitely does not cause any problems at all elsewhere in the programming process.

On a more serious note: if the these AI tools are the productivity boon people claim then we should start to see a revolution in software and office productivity in a few weeks. The upside of claims of revolutionary improvement is that it takes very little time for us to find out if they’re true or not.

Given the rate of adoption of these tools in the tech scene, whatever benefit they pose should start to become obvious by summer.

If the revolution doesn’t show up on time, hopefully some people will be introspective enough to realise that they fell for Automation and Anchoring biases, maybe with a heavy dollop of Anthropomorphism as a bonus.

And no, “I feel so much more productive, that’s proof” is nonsense. The proof is in the outcomes. How you feel while you’re doing it, while important for your own emotional wellbeing, is not an objective measurement of productivity benefits resulting in improved outcomes.

The human brain’s tendency towards anthropomorphism is so strong that you can, with no small amount of confidence, dismiss any claims that these AIs are capable of reasoning out of hand, unless those claims are backed by several solidly designed studies that have been replicated, with variations.

Y’know, backed by science.

For what it’s worth, my guess is that AI coding tools will end up being a 1-for-1 replacement for the open-source tools they’re leaching off.

As in, whenever you see a claim like “it would have taken me hours/days to build this chatbot, now I can do it in a fraction of the time”, what’s missing from the statement is that if it weren’t for the AI tool, they’d have used an open source chatbot framework and got there in the same amount of time, with fewer bugs, but more dependency issues.

For more of my writing on AI, check out my book The Intelligence Illusion: a practical guide to the business risks of Generative AI.

You can also find me on Mastodon and Twitter