4 ms·

I can understand companies not wanting to leak sensitive data or information, so not wanting their inputs used for training. But the individual level moral per

I can understand companies not wanting to leak sensitive data or information, so not wanting their inputs used for training. But the individual level moral perspective confuses me. You object to your own input being used for training "for free", but you're ok to use models which already slurped the data of millions of other people "for free"? I don't get it. Obviously this is going to be a controversial take on here (I am not blind to the sentiment), so, please help me understand. If you're a conscientious objector, why are you using the models in the first place?

greggoB 1mo ago
Private users also have sensitive information which, if leaked, could be used to their detriment (e.g training data getting hacked, etc). I don't see any distinction between companies and people when it comes to moral objections re training data - at least, if that's a thing, I'm unaware of it. I do agree with your last point re the hypocrisy.
dsign 1mo ago
There's something to this argument. I for one use an API proxy (Kagi Assistant) when I'm submitting anything remotely sensitive to an LLM. But for coding in my hobby project, I have marked "use my conversations for training" in the Claude app. I do it in the off chance that one day we will use stupid amounts of compute for personalized medicine and to fight cancer; code for scientific and technical computing is really complicated and only a tiny fraction of humans can produce it.
- munksbeer 29d ago
  If I understand your comment correctly (and highly likely I missed something), I am very similar. Using stuff at work the AI team tries their best to ensure we don't leak our internal data or code. But for my personal use, I'm totally fine with helping train the models. I'll be called naive, but I'm on the optimist side of the fence here. I hope, and expect, AI actually frees us up and delivers much improved lives for everyone. I know this goes against the current zeitgeist, but it is genuinely what I think will happen rather than the dystopia most people predict.

greggoB 1mo ago

Private users also have sensitive information which, if leaked, could be used to their detriment (e.g training data getting hacked, etc). I don't see any distinction between companies and people when it comes to moral objections re training data - at least, if that's a thing, I'm unaware of it. I do agree with your last point re the hypocrisy.

dsign 1mo ago

There's something to this argument. I for one use an API proxy (Kagi Assistant) when I'm submitting anything remotely sensitive to an LLM. But for coding in my hobby project, I have marked "use my conversations for training" in the Claude app. I do it in the off chance that one day we will use stupid amounts of compute for personalized medicine and to fight cancer; code for scientific and technical computing is really complicated and only a tiny fraction of humans can produce it.

munksbeer 29d ago
If I understand your comment correctly (and highly likely I missed something), I am very similar. Using stuff at work the AI team tries their best to ensure we don't leak our internal data or code. But for my personal use, I'm totally fine with helping train the models. I'll be called naive, but I'm on the optimist side of the fence here. I hope, and expect, AI actually frees us up and delivers much improved lives for everyone. I know this goes against the current zeitgeist, but it is genuinely what I think will happen rather than the dystopia most people predict.