I should stop reading HN altogether. I can't read one more comment about "How bad mathematicians reaction is" "how they were special and are not anymore and they can't cope"
Humanity is depressing.
Remote
Mathieu
@mathieu@fosstodon.org
mastodon 4.7.3Mainly a silent lurker. Interests include kernel dev and ... well that's pretty much it, it's vast enough.
Coffee ner^W enthusiast.
0 Followers
0 Following
21 Posts
Joined December 23, 2022
I went down a rabbit hole and scrolled through the reddits/forums of the auto proclaimed "accelerate" people (a term I just discovered mind you). Oh my I wasn't ready. It's a mix and match of different views but all of them are.... just weird, very very weird.
Where is your pride in the craft? Shattered right away when you were offered a magical Enter button ? blah. f that.
So I'm sure you know about the concept of mansplaining. Can we come up with a new term for the same situation but it's any human using an LLM?
This has to be one of the most infuriating feeling, especially when it's your own code -_-.
Seeing all over the places, comments like "we are just meat proxies now", "I press enter all day", "work is soulless"....
I don't think I've seen a bigger leopard ate my face kind of moment in a my life. whatever anyways let them have it.
Replying to
@david_chisnall@infosec.exchange Nah but that means you have to think. Nowadays you just throw wild ideas out and see what sticks, maybe the machine will "think" for you even. Bleh.
I can't wait for either the bubble to burst or the machine taking over. This inbetween is such a sad pathetic state.
Replying to
@seufz@troet.cafe @david_chisnall@infosec.exchange Yup, I'm a bit removed from all of this crazyness (by choice), I recently stumbled upon some "pro AI" communities and indeed they see "us" (wtf there is no us/them but okay) as technophobes trying to shut down everything out of fear. The world is wild.
Replying to
Replying to
... of domain knowledge. Because guess what? It will affect us, we will have to spend time reviewing this shit, we will have to fix it when it breaks. So yeah the shortcut cheat to mass producing code is pissing us off big times.
That's gatekeeping?
Replying to
@hyeyoo@social.kernel.org I wish I felt the excitement. I mostly feel disgust at the tech, and I do worry even though I don't think it'll succeed because the tech is great (it'll succeed because money).
And given all the excitement around I have to also question my sanity :)
Replying to
@david_chisnall Ah I remember being a student, going all in on metaprogramming, taking a step back and learning haskell just to be better at it. Coming back to TPMP sucked hard haha.
Good ol' days.
I tried. I really tried, I don't understand how can anyone take any joy or pride in writing plain English and then acting as a reviewer. It's pitiful, it's a miserable experience.
(also the constant back and forth and hand holding is just too much hassle)
Replying to
@hyeyoo@social.kernel.org The thing is while I've been thinking rationally about this, I've a feeling a large part of our peers haven't. My take is that the tech is "good enough" (see latest claude etc), that people will keep hyping it just enough that it will justify printing more money for it.
In the long run will we see the harsh reality of this (hard to fix bugs etc)? Sure! Will it be too late ? Pretty sure that's also a yes.
Anyway sorry my reasoning is a bit all over the place, it's just confusing times
Replying to
For a long time I couldn't pinpoint it to precise and concrete examples. I still can't but I have an analogy. You see how LLMs are pretty bad at writing, they just throw huge word salad at you, and, yes, you get the idea but looking at each sentence it's just gibberish.
Well guess what? Code they generate is the same.
Sure the model are now good enough that if you are lucky most of the time the code will do what is advertised, but I can't promise it because I can't wrap my head around it.