Agreed, but I can't imagine that everybody here on HN that's defending LLMs so... misguidedly and obviously ignoring observable reality... are financially interested in LLMs' success, can they?
I’m financially interested in Anthropic being successfull since it means their prices are more likely to go down, or for their models to get (even) better.
Honestly, if you don’t think it works for you, that’s fine with me. I just feel the dismissive attitude is weird since it’s so incredibly useful to me.
Given they're a VC backed company rushing to claim market share and apparently selling well below cost, it's not clear that that will be the case. Compare the ride share companies for one case where prices went up once they were successful.
Why is it weird, exactly? I don't write throwaway projects so the mostly one-off nature of LLM code generation is not useful to me. And I'm not the only one either.
If you can give examples of incredible usefulness then that can advance the discussion. Otherwise it's just us trying to out-shout each other, and I'm not interested in that.
It's a message board frequented by extremely tech-involved people. I'd expect the vast majority of people here to have some financial interest in LLMs - big-tech equity, direct employment, AI startup, AI influencer, or whatever.
Yeah, very likely. It's the new gold rush, and they are commanding wages that make me drool (and also make me want to howl in pain and agony and envy but hey, let's not mention the obvious, shall we?).
I always forget the name of that law but... it's hard to make somebody understand something if their salary depends on them not understanding it.
Orrrrr more simpler than imagining a vast conspiracy is that your observations just don't match theirs. If you're writing, say, C# with some esoteric libraries using CoPilot, it's easy to see it as glorified auto-complete that hallucinates to the point of being unusable because there's not enough training data. If you're using Claude with Aider to write a webpage using NextJS, you'll see it as a revolution in programming because of how much of that is on stack overflow. The other side of it is just how much the code needs to work, vs has to look good. If you're used to engineering the most beautifully organized code before shipping once, vs shipped some gross hacks you're ashamed of shipping, and the absolute quality of the code is secondary to it working and passing tests, then the generated code having an extra variable or crap that doesn't get used isn't as big an indightment of LLM-assisted programming that you believe it to be.
Why do you think your observable reality is the only one, and the correct one at that? Looking at your mindset, as well as the objections to the contrary (and their belief that they're correct), the truth is likely somewhere in-between the two extremes.
Where did I imply conspiracy? People have been known to turn a blind eye to criticism towards stuff they like ever since... forever.
The funny thing about the rest of your comment is that I'm in full agreement with you but somehow you decided that I'm an extremist. I'm not. I'm simply tired of people who make zero effort to prove their hypothesis and just call me arrogant or old / stuck in my ways, again with zero demonstration how LLMs "revolutionize" programming _exactly_.
And you have made more effort in that direction than most people I discussed this with, by the way. Thanks for that.
You said you couldn't imagine a conspiracy and I was responding to
that. As far as zero demonstration, simonw has a bunch of detailed
examples at:https://simonwillison.net/tags/ai-assisted-programming/ or maybe https://simonwillison.net/tags/claude-artifacts/, but
the proof is in the pudding, as they say, so setting aside some time and
$20 to install Aider and get it working w/ Claude, and then building a
web app is the best way to experience either the second coming, or an
overhyped let down. (or somewhere in the middle.)
Still, I don't think it's a hypothesis that most are operating under, but
a lived experience that either it works for them or it does not. Just the
other day I used ChatGPT to write me a program to split a file into
chunks
along a delimiter. Something I could absolutely do, in at least a
half-dozen languages, but writing that program myself would have
distracted me from the actual task at hand, so I had the LLM do it. It's
a trivial script, but the point is I didn't have to break my
concentration on the other task to get that done. Again, I absolutely
could have done it myself, but that would have been a total distraction.
https://chatgpt.com/share/67655615-cc44-8009-88c3-5a241d083a...
On a side project I'm working on, I said "now add a button to reset the camera view" (literally, aider can take voice input). we're not quite at the scene from star trek where scottie talks into the mac to try and make transparnet aluminum, but we're not that far off! The LLM went and added the button, wired it into a function call that called into the rendering engine and reset the view. Again, I very much could have done that myself, but it would have taken me longer just to flip through the files involved and type out the same thing. It's not just the time saved, which, I didn't have a stopwatch and a screen recorder, but apart from the time, it's not having to drop my thinking into that frame of reference, so I can think more deeply about the other problems to be solved. Sort of why ceo isn't an IC and why IC's aren't supposed to manage.
Detractors will point out that it must be a toy program, and that it won't work on a million line code base, and maybe it won't, but there's just so much that LLMs can do, as they exist now, that it's not hyperbole to say it's redefined programming, for those specific use cases. But where that use case is "build a web app", I don't know about you, but I use a lot of web apps these days.
These are the kind of the informed takes that I love. Thank you.
> Detractors will point out that it must be a toy program, and that it won't work on a million line code base, and maybe it won't
You might call me a detractor. I think of myself as being informed and feeling the need to point out where do LLMs end and programmers begin because apparently even on an educated forum like HN people shill and exaggerate all the time. That was the reason for most of my comments in this thread.
And it’s the short wrong answer. I get absolutely zero benefit if you or anyone else uses it. This is not for you or I, it’s for young people who don’t know what to do. I’m a person who uses a few LLM’s, and I do so because I see the risks and want to participate rather than have change imposed on me.
