I'm not even talking about holographic game consoles or FTL travel. Why can't we finally have a good container orchestrator that's not kubernetes for example?
Easy, we can't because the LLMs know a lot about kubernetes but wouldn't know anything about a new orchestrator, so they'll always suggest k8s, and the button-pushers will just say "yes, do it, no mistakes". Maybe if Anthropic decides they need to create a new orchestrator, they would bias Claude heavily enough for it to actually get used, but that's about the only way I see.
"Why don't we have any new programming languages" or "Why don't we have any new web frameworks" or "Why don't we have any new operating systems" or "Why don't we have any new container orchestrators" might be the same kinds of questions as "why is the design for washing machines still basically the same as it was 50 years ago?" or "why are we still using English, where are all the new conlangs to replace it?" "why are we still using email/SMTP instead of inventing a better system" "why haven't we invented a new calendar?"
All it will lead to is yet more people that are out of a job. Right now at least a driving license (more so if it is a commercial license) is a source of income to many.
What's the market demand? I'm confident there are plenty of alternatives being churned out today but nobody's looking or asking for any. If only because they settled on k8s, they have it on their CV, they have certifications, and switching is a huge investment and risk.
Right now to me it feels like the innovation is in using LLMs to (finally?) clean up existing code, find vulnerabilities, do maintenance, optimize, and if necessary rewrite.
Where before humans learned and built better tools, now you keep paying forever to get re-trained models. What makes it a regressive feedback loop is that the people who used to learn and push the frontier are now using AI to repackage others' work as their own. Eventually what will be left to train on but copies of copies of copies?