Personally, I don’t think that LLMs are useful for documentation. I feel that it only helps you have documentation and won’t help with actual useful documentation. Just a single hallucinated aspect of a documentation renders it almost useless.

In my experience it only generates adequate documentation when the software is simple and already self explanatory. It does not however help with actual use cases that might need to be documented because they are somehow difficult.

Software developers that are not interested in providing documentation will probably also not be interested in checking whether the documentation slopped out by an LLM is actually correct.

Some Proton Employee really contacted me here on lemmy to basically say that Proton could never do anything malicious, because they are owned by a non profit. Then how the heck can OpenAI be this scummy when they are also owned by a non profit? (See my Comment History)
Protons PR Team is so scummy. They spread misinformation about themselves in public forums and pay Content Creators to say incorrect things about online privacy to sell products to customers that don’t need those products.

Interesting idea. Cut enough heads of the Hydra until the heads are so small, that they don’t pose any threat.

Okay, so? OpenAI is also controlled by a Non-Profit (Similar Structure to Proton, I suppose). That doesn’t stop them from being scummy. Please explain your point.

Doesn’t work like that. It’s more like fighting a Hydra. If a Billionaire worth $50B dies, like 10 Children are gonna inherit $5B each. So we would have even more Billionaires afterwards.

Nothing is stopping Proton from doing the same thing next week. And seeing how many people lock themselves in to Proton (by using all their services, Apple style), they have a strong incentive to also do some “restructuring” and spike prices.

Second Mistake was using Windows in 2026

Why is Jetbrains YouTrack in the American circle?