I would like general purpose open source AI models to be legal, even if it is possible to use them for harmful purposes. Just like open source cryptography, open source operating systems, et cetera.
If a product's primary purpose is crime, fine, ban it. But AI models should be treated just like other technologies, from email to the web to the computer itself.
Comments
Is it your opinion that since it’s possible to subvert it they should abandon the attempt?
I would like general purpose open source AI models to be legal, even if it is possible to use them for harmful purposes. Just like open source cryptography, open source operating systems, et cetera.
If a product's primary purpose is crime, fine, ban it. But AI models should be treated just like other technologies, from email to the web to the computer itself.