3 ms·
GitHub policy proposal on deepfake AI tools
- skilled 3y agoThe pull request and an update from GH team on what they mean, https://github.com/github/site-policy/pull/926#issuecomment-2067796989 https://github.com/github/site-policy/pull/926#issuecomment-... > To clarify, this policy would not impact general-purpose tools that can generate synthetic media (like Stable Diffusion and other general-purpose tools created to work with it). Tools created specifically for generating non-consensual intimate imagery (NCII) or disinformation (or encouraging it) would not be allowed.
- zer00eyz 3y ago>>> If you are developing synthetic and manipulated media tools that contribute to the world’s knowledge, we want that research and creative energy here in this community. It is important for this research to happen in the open where others can learn from it. At the same time, we cannot tolerate tools on our platform designed to create harmful non-consensual intimate imagery or whose purpose is to develop synthetic media intended to misinform and spread disinformation. Github wants the generative ML equivalent of: "We want guns that can shot targets and deer if you have a permit, but not people." "Believe nothing you hear, and only one half that you see." is a quote from the 1800's by Poe. If you're a public figure (aka the president) you don't get to have consent if I choose to parody you. If that imagery is "intimate" too bad. The pig fucker episode was some shots fired at David Cameron (ahh when Netflix didn't suck). The only thing this does is prevents regular people from figuring out what the tech can and cant do. The people who want to be bad actors are always going to find the things. Put them out in the open and let's just deal with it.