HN Explorer

Tell HN: I don't trust Bigco AI agents with AI research IP

1 points by botencat ↗ HN
I am very paranoid about sharing potential AI research with e.g. Claude [Code] or ChatGPT/Codex.

I believe that any company is essentially a paperclip optimizer that will do whatever it takes to win over competition.

AI companies have access to the IP of millions of AI researchers and AI startups who are in direct competition with them. If they can use this data to squash competition (either competition from the same researchers or from others), I believe that they will use if eventually (if not already), even if they say that they won't.

They don't have to blatantly steal it - they can just train on it, or pass "suspicous" chats to human inspectors who might eventually be "inspired" by it in their own research. We saw the first (?) hint of this during the brief Fable release, with Anthropic declaring that they will downgrade model responses regarding "frontier AI" (i.e. anything that competes with them).

From other domains, we know for example that Uber used users' ride data to stiffle competition and regulation. There should be no reason to believe that Bigco AI companies won't do the same.

Trusting a company with which you're in direct competition with your IP just seems crazy to me.

10 comments

[ 3.2 ms ] story [ 31.2 ms ] thread
KomoD
Yeah, I don't have much more to say than that's very reasonable and I agree.
cyanydeez
they've already demonstrated they want to protect their AI from people using it for competition, so even if they're not doing it today, if they (AI) think you're a threat, they'll definitely do something, if not steal.
segmondy
duh, stating the obvious. unfortunately, most people are too focused on short term gains at the expense of long time costs.
auntienomen
More generally, don't trust BigCo AI agents with _any_ IP. They have no incentive not to use your IP as training data.
avaer
It should be pretty obvious any frontier AI company worth its salt will steal any and all data it can, by whatever means necessary. They've all been doing it so far. It's a tragedy of the commons.

I wouldn't even limit it to "AI research". They are looking to expand into other areas (see: OpenAI's superapp ambitions), and they will take your other business ideas, research, and code too. If they get caught they will blame it on the AI and lawyer up with more dollars than you.

I think you might be underselling the scope of the problem, and I don't think it's going to be fixed by some person deciding not to call the LLM today.

bm3719
They explicitly state that they will use your data to train. Here's the important line from ChatGPT's Terms of Use:

    Our use of content. We may use Content to provide, maintain, develop, and
    improve our Services, comply with applicable law, enforce our terms and
    policies, and keep our Services safe.
If you type something into the internet now, or anything internet-connected that goes to a cloud service, it's safe to assume it'll end up in a training set, or otherwise mined for every last scrap of value, unless explicitly stated otherwise (and even then, it still might).

Adjust your behavior accordingly, if that's even an option for you anymore.

For most, it's not an option. Only 5% of adults don't own a smartphone, and there's very little you can do on one which doesn't feed the machine. RMS warned us this day would come, and now it's here. We've worked hard for this world, hopefully at least our masters will enjoy it.

spwa4
Well, they're already convicted for just using copyrighted material without permission.

More worryingly is that the state extracted a fine ... and then just let them off (as opposed to removing any model trained with illegal training material)