Not even 10 percent is. Good lord. Go outside and touch grass.

Not really. At some point the technical debt accumulates and the only option is to trash it all and start over.

The only party that profits here are the cloud token providers.

That's a big ask. No thanks.

LLMs have a limit to how deep they can understand and refactor architectural issues.

That limit is far, far lower than a human's.

Also, you're confusing "AI" and LLMs.

This has been true for the last 60 years.

Generating more technical debt is never a difficult problem in software development.

Good lord!

I hope somebody is documenting all these bombastic LLM-related public statements. They're going to be a neverending source of cringe and laughs for the next 50 years.

(I've heard one C-level dude say with a straight face that LLMs were a "more significant invention than writing".)

No it won't. Natural language manipulation is never a bottleneck in STEM.

Code generation is a very silly way of using LLMs. They're not even good at it.

It matches with Reddit posts that have statistically similar words and starts generating the next statistically likely token.

There's laws and shit about that. You can't just immigrate to a country illegally.

They literally are. (If by "all this" you mean the subscription future bait-and-switch plans.)

(Real mathematics problems, not American-style ""math"".)

Python is the antistandard for package management. Or maybe even the eldritch horror of package management.

This shows that AI cloud consumption is just a conspicuous consumption status symbol, nobody knows why they need cloud AI or what problem they are even solving.