That’s not by design, That’s a failure state known as overfitting. This happens when you have the same thing repeated in your training set too many times, and it effectively gets “burned in”. The whole point of generative AI is to create new things, why would you ruin a perfectly good system to accomplish the same thing as clicking “save as”?
It was Midjourney that had the Afghan Girl problem, BTW, not GPTs.
I guess I want free art, you should too. There is nothing wrong with wanting a tool to help people better communicate, inspire, create, and connect with each other in ways they may not have been able to before.
We’re all standing on the shoulders of giants. We learn from each other, and humanity is at its best when we can all share in our advancements. Calling this stealing is self-serving, manipulative rhetoric that unjustly vilifies people and misrepresents the reality of how these models work.
It took us 100,000 years to get from cave drawings to Leonard Da Vinci. This is just another step, like artists who used Camera Obscura in the past. It’s important to remember that early man was as smart as we are, they just lacked the interconnectivity that we have.
I don’t think there is some loophole that is against the spirit of the law, it works like this on purpose. In the US, fair use balances the interests of copyright holders with the public’s right to access and use information. There are rights people can maintain over their work, and the rights they do not maintain have always been to the benefit of self-expression and discussion. We shouldn’t be trying to make that any worse.
I disagree with that article 100%. They’re missing the spirit of copyright law which is to protect artists that create original work. They already try to protect musicians and do this in music with sampling. But instead of one artist stealing parts of your song or the spirit of the song, they’re coming after visual art from every angle because it’s a computer and like 10,000 people doing that. We should be able to copyright our work and have to agree for it to be used, the end. It’s ridiculous and greedy of these huge companies to do anything else.
Just to be clear, I don’t give a shit about corporations, but what you want will hurt all artists and give corporations the unprecedented legal tools to take down anything they don’t feel like having around.
What part are you referencing here:
Also, how fucked up is it that they liken it to the little guy competing with big companies, that’s not a thing. If they used public domain art, I don’t think any of this would be an issue. Do we own our own voice, imagination, likeness? I say yeah, we do. The corporations shouldn’t have the rights to those.
Also remember that AI training isn’t only for mega-corporations. We can already train open source models, we shouldn’t put up barriers that only benefit the ultra-wealthy. If we weaken fair use, we hand corporations a monopoly of a public technology by making it prohibitively expensive to for regular people to keep up. Mega corporations already own datasets, and have the money to buy more. And that’s before they make users sign predatory ToS allowing them exclusive access to user data, effectively selling our own data back to us. Regular people, who could have had access to a competitive, corporate-independent tool for creativity, education, entertainment, and social mobility, would instead be left worse off and with fewer rights than where they started.
Luckily, AI is a public technology. That’s why they’re already trying their hand at regulatory capture. And they might just get it. Just like they’re trying to destroy encryption. Support open source development, It’s our only chance. Their AI will never work for us. John Carmack put it best.