@oblomov @mcc @mark @gsuberland How, though? Let's...

@oblomov @mcc @mark @gsuberland How, though?

Let's say that tomorrow, a judge rules that GPT-4 is not covered by copyright. What has actually changed? OpenAI isn't compelled to share it with anyone, and it's too big for anyone except large and wealthy corporations to actually do anything with.

Sure, you couldn't get sued if you got a bittorrent of it somehow. But you're not getting a bittorrent of a 1.76 trillion parameter neural network anyway.

Like 3 April at 21:28 | Wall-to-wall | Open on hachyderm.io

3 comments

Graham Spookyland🎃/Polynomial replied to datarama

@datarama @oblomov @mcc @mark and you sure as shit can't afford a whole rack of H200 cards to make use of it, even if you and all your friends pitch in. it's only useful with people who have the capital to wield it.

3 April at 21:31 | Open on chaos.social

crzwdjk ✅ replied to datarama

@datarama @oblomov @mcc @mark @gsuberland 1.76 trillion parameters is about a hard drive's worth of data, no?

4 April at 0:23 | Open on mastodon.social

datarama replied to crzwdjk ✅

@crzwdjk @oblomov @mcc @mark @gsuberland It is, but that's *still* beside the point. You can't actually do anything with it unless you have the resources of a large corporation.

And my other point was that just because it isn't copyrighted, they can still keep it secret.

4 April at 5:33 | Open on hachyderm.io