Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

The AI Vampire

steve-yegge.medium.com

Post reply on HN
Print view View original

31–40 of 107 posts

Re: The AI Vampire

#31

Post by WithinReason »

augment_me wrote: post #29
dwedge wrote: post #13

Every time I say I don't see the productivity boost from AI, people always say I'm using the wrong tool, or the wrong model. I use Claude with Sonnet, Zed with either Claude Sonnet 4 or Opus 4.6, Gemini, and ChatGPT 5.2. I use these tools daily and I just don't see it. The vampire in the room, for me, seems to be feeling like I'm the only person in the room that doesn't believe the hype. Or should I say, being in roo…

I am with you on this, and you can't win, because as soon as you voice this opinion you get overwhelmed with "you dont have the sauce/prompt" opinions which hold an inherent fallacy because they assume you are solving the same problems as them. I work in GPU programming, so there is no way in hell that JavaScript tools and database wrapper tasks can be on equal terms with generating for example Blackwell tcgen05 warp…

In my experience LLMs are useless for GPU compute code, just not enough in the training set.

Re: The AI Vampire

#32

Post by benreesman »

I am a long time fan of Steve Yegge but he's too much part of the groupthink at this point.

You can't win with Claude Code. I understand that his API key isn't on the PID controller, so he gets a less bad deal, but he's still breaking even with some gee whizz factor.

Agents are like people on a long enough timeline: they will eventually do the lazy thing. But this happens in minutes not years.

If you don't have them on tracks made of iron, you are on a sugar high that will crash.

Formal methods, zero sorry, or it's another bounty for the vibecode cleanup guys.

Re: The AI Vampire

#33

Post by exfalso »

dwedge wrote: post #13

Every time I say I don't see the productivity boost from AI, people always say I'm using the wrong tool, or the wrong model. I use Claude with Sonnet, Zed with either Claude Sonnet 4 or Opus 4.6, Gemini, and ChatGPT 5.2. I use these tools daily and I just don't see it. The vampire in the room, for me, seems to be feeling like I'm the only person in the room that doesn't believe the hype. Or should I say, being in roo…

Exact same experience.

Here's what I find Claude Code (Opus) useful for:

1. Copy-pasting existing working code with small variations. If the intended variation is bigger then it fails to bring productivity gains, because it's almost universally wrong.

2. Exploring unknown code bases. Previously I had to curse my way through code reading sessions, now I can find information easily.

3. Google Search++, e.g. for deciding on tech choices. Needs a lot of hand holding though.

... that's it? Any time I tried doing anything more complex I ended up scrapping the "code" it wrote. It always looked nice though.

Re: The AI Vampire

#34

Post by bsaul »

mbgerring wrote: post #25

This is a good time to repeat that software engineers need a union. We needed this ten years ago, and we need it a lot more now.

As a european, yes please America, get a union. Get 2 even. You're going too fast, you're way too successful, we can't keep up.

So yes, please adopt our work ethic and legal framework. It's going to help us tremendously.

Re: The AI Vampire

#35

Post by Davidzheng »

dwedge wrote: post #13

Every time I say I don't see the productivity boost from AI, people always say I'm using the wrong tool, or the wrong model. I use Claude with Sonnet, Zed with either Claude Sonnet 4 or Opus 4.6, Gemini, and ChatGPT 5.2. I use these tools daily and I just don't see it. The vampire in the room, for me, seems to be feeling like I'm the only person in the room that doesn't believe the hype. Or should I say, being in roo…

I don't understand what including the time of "4 years" does for your arguments here. I don't think anyone is arguing that the usefulness of these AIs for real projects started at GPT 3.5/4. Do you think the capabilities of current AIs are approximately the same as GPT 3.5/4 4 years ago (actually I think SOTA 4 years ago today might have been LaMDA... as GPT 3.5 wasn't out yet)?

Re: The AI Vampire

#37

Post by mrkeen »

mbgerring wrote: post #25

This is a good time to repeat that software engineers need a union. We needed this ten years ago, and we need it a lot more now.

What concrete interests would you like such a union to protect?

Should a strike happen if devs are told to use Claude, or should a strike happen if devs aren't given access to Claude?

Re: The AI Vampire

#38

Post by davedx »

mjr00 wrote: post #9

I think Yegge hit the nail on the head: he has an addiction. Opus 4.5 is awesome but the type of stuff Yegge has been saying lately has been... questionable, to say the least. The kids call it getting "one-shotted by AI". Using an AI coding assistant should not be causing a person this much distress. A lot of smart people think they're "too smart" to get addicted. Plenty of tales of booksmart people who tried heroin…

I'm a recovering alcoholic. One thing I learned from therapists etc. along the way is that there are certain personality types with high intelligence, and also higher sensitivity to other things, like noise, emotional challenges, and addictive/compulsive behaviour.

It does not surprise me at all that software engineers are falling into an addiction trap with AI.

Re: The AI Vampire

#39

Post by augment_me »

WithinReason wrote: post #31
augment_me wrote: post #29

Earlier quoted context omitted.

I am with you on this, and you can't win, because as soon as you voice this opinion you get overwhelmed with "you dont have the sauce/prompt" opinions which hold an inherent fallacy because they assume you are solving the same problems as them. I work in GPU programming, so there is no way in hell that JavaScript tools and database wrapper tasks can be on equal terms with generating for example Blackwell tcgen05 warp…

In my experience LLMs are useless for GPU compute code, just not enough in the training set.

Yeah, the argument here is that once you say this, people will say "you just dont know how to prompt, i pass the PTX docs together with NSight output and my kernel into my agent and run an evaluation harness and beat cuBLAS". And then it turns out that they are making a GEMM on Ampere/Hopper which is an in-distribution problem for the LLMs.

It's the idea/mindset that since you are working on something where the tool has a good distribution, its a skill issue or mindset problem for everyone else who is not getting value from the tool.

Re: The AI Vampire

#40

Post by Terr_ »

dwedge wrote: post #13

Every time I say I don't see the productivity boost from AI, people always say I'm using the wrong tool, or the wrong model. I use Claude with Sonnet, Zed with either Claude Sonnet 4 or Opus 4.6, Gemini, and ChatGPT 5.2. I use these tools daily and I just don't see it. The vampire in the room, for me, seems to be feeling like I'm the only person in the room that doesn't believe the hype. Or should I say, being in roo…

I'm an AI hipster, because I was confusing engagement for productivity before it was cool. :P

TFA mentions the slot machine aspect, but I think there are additional facets: The AI Junior Dev creates a kind of parasocial relationship and a sense of punctuated progress. I may still not have finished with X, but I can remember more "stuff" happening in the day, so it must've been more productive, right?

Contrast this to the archetypal "an idea for fixing the algorithm came to me in the shower."

Post reply on HN