(response written without the use of AI)

https://mastodon.social/@Daojoan/117125213303488912

LLM is a power-amplification tool, like a hammer or a backhoe.

With a skilled operator: A very good thing indeed.

Operated by a careless or clueless person: A danger to life, limb and property.

And, to me, that is very different from an LLM patch. You have put in the effort. As with everyone, there is a limit to what your effort will achieve that is bracketed by your knowledge, but you’re asking for help after trying, and you’re explicit about where the edges of your knowledge are.

I love getting patches like that because they come from someone who has identified a problem and the kinds of things that they get wrong are predictable and so review is fairly easy. They generally fall into two categories: either they’re nearly right and a small tweak is all that’s needed to land them, or there’s some fundamental misunderstanding that means the code is useless as is, but serves as a very detailed bug report for something people have actually encountered as a problem in real use. And these are both things that have value to me and so make me want to put in the remaining effort to make sure that we have something that benefits us both.

And, if someone who doesn’t think they’re a programmer (you wrote some code and sent it to a real project, you’re a programmer, sorry to be the one to break this to you) put in the effort, that’s also a strong signal to me that this is a fix that actually matters to users.

If you hadn’t bothered to try to understand the codebase at all and just sent an LLM patch, it’s more work to review because LLMs have very different (and far more complex) failure modes than inexperienced programmers. The effort to say ‘Claude, do the thing’ is so low that I don’t know if the person who did it even uses the project or if they were just looking for something to pad their CV, so I have no idea if this is a change anyone wants. And if you don’t understand what the change you’re making does, you’re putting 100% of the cognitive load on me.

Some of the most useful contributions I’ve had to various F/OSS projects have not even been code. Detailed bug reports with test cases that let me reproduce them are incredibly valuable.

For each of the complaint, a diligent, caring LLM user could actually use LLM to reduce the burden on others by having the LLM support their thinking.

If you write something, have the LLM take a pass at pointing out lack of clarity, duplication, style problems.

If you write code, have an LLM point out where you overlooked an edge case, missed clarity in a comment, left a bug…

Perhaps it's not so much the tool as the user that is at fault?

"Someone always has to do the work that a message gets understood. It can be either the sender or the receiver. If you are the sender, do you want to leave it up to the receiver, who usually has less incentive to do the work?"

Now this is my north star when writing emails, bulletins, news, etc. You can use LLM to help you with this process (not saying you should use it) but you cannot outsource the provess to an LLM or you are actually moving the work back to the receiver.

Playing games of Go now - a deeply complex game requiring thought and analysis to play well - most players play insanely fast, rarely speak and just do not care in the slightest that they are playing another human.

It is very chilling how deep the lack of humanity can be.

And people have been doing such things a lot without LLMs, too.

@david_chisnall

Digging through an artbook, writing a few sentences, stick figures, hero forge, etc is all the client needs to do. That’s not much work at all, but it matters.

“Look at this post [A]I have written to make me look good.”

I greatly improved my quality of life by deleting that app.

That which we find annoying in others - they find annoying in us.

Really easy test: did the critics take linear algebra. If not, they aren’t talking about AI