There's an old and famous quote around computers not having morals, and thus computers requiring people on top of them that can steward systems and be hold accountable.

Besides morals, you'll always need people in the interface between the computers and the world. Maybe they won't write too much code, but they'll need to specify and verify behaviours anyway.

wishful thinking. Governments around the world are relying on algorithms to decide who can have welfare or who is cheating on taxes.

With the rise of AI it will only get worse.

These algorithms were designed and verified by humans.

Non deterministic software will never produce 100% of the time perfect results. So we will need humans with knowledge to verify up to when an other technology will be able to build enough trust to avoid human reviews. With 2000 line very precise prompts, Claude code finishes saying the code is ready for production and there is still anomalies and bugs that it should have catched but didn’t. Good luck trusting 100% of your business on such technologies.

It's not that good, and almost always more complex than the essence a senior person will boil thing down to.