It doesn't have to replace people on a one-for-one basis to cause job losses. Let's say LLMs make your developers 50% more efficient. Doesn't it stand to reason you can lay off the lowest performing 33% and get the same amount (or more) of work?

No, it does not stand, because you think linearly. Companies can't simply drop 33% of employees because there is competition. If competition uses both humans and AI they will get more value from both. No AI has sufficient autonomy or capability to be held accountable for its mistakes.

There is less upswing in reducing costs than in increasing profits. Companies want to increase profits actually, not just reduce costs which will be eaten away by competition. In a world where everyone has the same AIs, human still make the difference.

It also means that with lower costs your service becomes more attractive and maybe attracts more customers, so might even grow the number of workers.

This is known as Jevons Paradox https://en.wikipedia.org/wiki/Jevons_paradox

there hasn’t been a single study that concludes any benefits to AI yet.

Either it’s a cover for something or people are a bit too overzealous to believe in gains that haven’t materialised yet.

Have you ever been at a company where the limiting factor was finding stuff to build? I've never seen one personally. If there's any productivity increase, they'll just build even more stuff.

(And that's if we agree about a 50% increase I'd say 5% is already generous)