3 ms·

Well llms are imperfect tools that provide a lot of value (they offer leverage in form of agents to put it in Naval Ravikant words). They are here to stay.

by itanium 17d ago

verdverm 17d ago
> Fundamentally I don’t think we have a choice about riding on the AI technology train. It’s a wild ride and I just hope we’ll get through it OK. > I don’t like them. They talk to me in this grating LLM-voice, an uncanny valley of talking to a real human. They confidently bullshit me - often giving me useful, helpful answers. But also just making stuff up with the same assurance - and with only a veneer of fake remorse when I call them out on it. I think what he needs is a personal fine-tune, something I suspect will become more prominent when we start getting multi-tenant LoRA and can use fine-tunes on a per-token cost basis. A co-worker commented that he thought the models performed worse after being "embarrassed," suspect it is an artifact of how some conversations progress in the training data.
- skeledrew 17d ago
  I've gotten better results both after "shouting at" and praising Claude, though the former causes more token burn because it starts questioning asking permission for everything after, which gets really annoying. Even a session clear doesn't get it completely back to baseline; only after some time does it return to "its old self".
  - verdverm 17d ago
    With opencode, I just start a new session and it's the same blank slate, there's no lingering of prior state Does /clear in claude not do what I intuit it to?
    - skeledrew 17d ago
      It does, but I feel like something is retained somewhere. There's already auto-memory, and the web chat has a persistent memory feature. It wouldn't surprise me if there's some hidden state that leads to per project or even per user auto-customization over time.
ashkankiani 17d ago
He says this in the post almost word-for-word. Are you agreeing with him? The opening "Well" confuses me on what tone you're going with.
happytoexplain 17d ago
This comment doesn't make sense as a response to anything the author said. As an aside: > they offer leverage in form of agents to put it in Naval Ravikant words This is a confusing use of attribution. It's sort of like saying, "as Naval Ravikant famously said, AI can be very useful."
- itanium 17d ago
  [dead]
xdavidliu 17d ago
> offer leverage in form of agents i would expect that when you cite someone by name, the sentence being cited should at least be significant and meaningful to the slightest degree
dominotw 17d ago
you thought this was some sort of insightful rebuttal that you uniquely thought of?

Well llms are imperfect tools that provide a lot of value (they offer leverage in form of agents to put it in Naval Ravikant words). They are here to stay.

by itanium 17d ago

Well llms are imperfect tools that provide a lot of value (they offer leverage in form of agents to put it in Naval Ravikant words). They are here to stay.

verdverm 17d ago

> Fundamentally I don’t think we have a choice about riding on the AI technology train. It’s a wild ride and I just hope we’ll get through it OK. > I don’t like them. They talk to me in this grating LLM-voice, an uncanny valley of talking to a real human. They confidently bullshit me - often giving me useful, helpful answers. But also just making stuff up with the same assurance - and with only a veneer of fake remorse when I call them out on it. I think what he needs is a personal fine-tune, something I suspect will become more prominent when we start getting multi-tenant LoRA and can use fine-tunes on a per-token cost basis. A co-worker commented that he thought the models performed worse after being "embarrassed," suspect it is an artifact of how some conversations progress in the training data.

skeledrew 17d ago
I've gotten better results both after "shouting at" and praising Claude, though the former causes more token burn because it starts questioning asking permission for everything after, which gets really annoying. Even a session clear doesn't get it completely back to baseline; only after some time does it return to "its old self".
- verdverm 17d ago
  With opencode, I just start a new session and it's the same blank slate, there's no lingering of prior state Does /clear in claude not do what I intuit it to?
  - skeledrew 17d ago
    It does, but I feel like something is retained somewhere. There's already auto-memory, and the web chat has a persistent memory feature. It wouldn't surprise me if there's some hidden state that leads to per project or even per user auto-customization over time.

ashkankiani 17d ago

He says this in the post almost word-for-word. Are you agreeing with him? The opening "Well" confuses me on what tone you're going with.

happytoexplain 17d ago

This comment doesn't make sense as a response to anything the author said. As an aside: > they offer leverage in form of agents to put it in Naval Ravikant words This is a confusing use of attribution. It's sort of like saying, "as Naval Ravikant famously said, AI can be very useful."

itanium 17d ago
[dead]

xdavidliu 17d ago

> offer leverage in form of agents i would expect that when you cite someone by name, the sentence being cited should at least be significant and meaningful to the slightest degree

dominotw 17d ago

you thought this was some sort of insightful rebuttal that you uniquely thought of?