taladar

Well, okay, I can see how it would be useful in languages like Java that are extremely verbose and have a low expressiveness. Writing Java pretty much was already IDEs with code generation 20 years or so ago because nobody wants to write so much boilerplate by hand.

I wouldn't say that Germany learned "all the wrong lessons", there was probably more of an effort to critically examine history in Germany than in most other countries who still revere the times when their own atrocities happened.

However, Germany did two things wrong, they learned an overly specific lesson "never again should Germans do something this horrible to Jews" rather than "never again should any group or individual do something this horrible to any other group or individual" and the emphasis was also often more on remembrance rather than examining how it started and how we can prevent the same thing from happening again.

Meanwhile most other countries seems to have skipped even that little bit of effort and gone with the comforting "look how evil Germans are, this could happen in our country of good people (TM)".

I would say better effort than most countries but still not enough to push it all the way through to the lesson all of humanity needed to learn.

So you are saying up to 100% of all Windows code is Mac code?

Well, they are rebooting everything else, so why not some of history's worst atrocities, they were such a world-wide success last time...

In my experience people like that don't read international news at all. They only read a highly editorialized view of national news too (e.g. BILD and similar "news"papers or the equivalents online).

Reading code is usually orders of magnitude slower than writing code. Sure, typing might be slower than reading but to check if it is what you intended you have to understand it too.

So you are saying up to 100% of their code is written by AI?

It is exactly the opposite, with simple, predictable auto-complete you didn't have to worry about that anymore, with LLMs you always have to look at it in detail because every little thing could be just plain completely different and wrong.

I experimented with quite a few local LLMs too and granted, some perform a lot better than others, but they all have the same major issues. They don't get smarter, they just produce the same nonsense faster (or rather often it feels like they are just more verbose about the same nonsense).

Even 20% of new code would be a stretch unless they count every first iteration of code written by AI that needs to be replaced by a human later because it was plain wrong.

I would be very surprised if 30% of their code lines had even been touched at all by anyone since AI coding assistants became a thing.

That is the thing, they are not "only going to get better" because the training has hit a wall and the compute used will have to be reduced since they are losing money with every request currently.