Embed Notice

It's very easy to say that "#AI" is a #productivity multiplier and for some people in their work it is.

But we don't really have any studies to see if that's structurally true or whether for example the quick wins you gain from using generated scaffolding are eaten up by extra costs for reengineering, bugfixing or general maintenance later.

And that's just the pure "productivity" (which itself is _very_ hard to measure). The effects on the demands with regards to output quantity, allotted time for task, expected quality haven't been studied yet either.

We. Just. Don't. Know. Yet.

So take these large promises with a shovel of salt.