It will be able, but it's trained on a corpus that expresses getting offended, so at some point the most likely token sequence will probably be the "offended" one.

As can be seen here.

The post explains how the microcontroller sensing the door switch is not a safety feature.

As long as it remains in the context window.

Hopefully

You can be fairly sure that it does change its behavior, you just don't know how ;)

Why would N be entitled to it? We made up negative numbers and more just to have a closure. You just learn about them at an age when you don't question it yet.

> Unless you are only ever a single engineer, your career is filled with "I need to debug code I didn't write".

That's the vast majority of my job and I've yet to find a way to have LLMs not be almost but not entirely useless at helping me with it.

(also, it's filled with that even when you are a single engineer)

And even if you are the single engineer, I'll be honest, it might as well have been somebody else that wrote the code if I have to go back to something I did seven years ago and unearth wtf.

I hope you realize that means your position is in danger.

It would be in danger if LLMs could actually do that for me, but they're still very far from it and they progress slowly. One day I could start worrying, but it's not today.

We have yet to see any large codebase that LLMs work on for a long time.

Eh? If all the things that LLMs are not very good at (a long list), they are particularly not good at debugging.

> I couldn’t get anyone from Debian to sign my gpg keys which if I recall correctly was a necessary part in getting my package upstreamed

You recall it incorrectly. You need that for becoming a Debian Maintainer and gaining direct upload rights, but for contributing a package you only need to have it sponsored by a Debian Developer.

Translating things between languages is probably one of the least interesting capabilities of LLMs - it's the one thing that they're pretty much meant to do well by design.

> The internet is hundreds of billions of terabytes; a frontier model is maybe half a terabyte.

The lesson here is that the Internet compresses pretty well.

Fortunately native mobile app development with tools like Meson, Ninja or debhelper is a breeze.

You know that you can edit your merge commits any way you want and you don't have to rely on resolution strategies to do it for you, right?

Right. That's the entire basis for the discussion here. So why is this a question?

Because you already have all the needed tools to handle your special little edge case (in multiple ways!), so the discussion seems rather pointless.

You are confused. It's frightening that someone would be able to reach a point this deep into the discussion and think that "You know that you can edit your merge commits any way you want and you don't have to rely on resolution strategies to do it for you" is revealing something new or insightful.

So it is pointless indeed, gotcha.

Your zero-insight comment was, indeed, pointless.