The US isn’t introducing bias into the models, they’re introducing bias into the internet. The AI models just reflect the larger problem, which is, bias is bad.
Any point of origin is going to introduce its own biases, whether they intend to or not. The solution is to have AI development be decentralized, and have as many points of origin as possible. After introduction most of the bias seen will be a result of the majority of the information it is encycloping.
I don’t know what point you’re trying to make. The data set used is world wide.
My point is it’s an American technology, so the Americans get to introduce whatever bias they want. Yes, that bias is controllable.
Not to sound dismissive but if you want one that doesn’t have American bias, write it yourself. Nothing stopping you.
The US isn’t introducing bias into the models, they’re introducing bias into the internet. The AI models just reflect the larger problem, which is, bias is bad.
Any point of origin is going to introduce its own biases, whether they intend to or not. The solution is to have AI development be decentralized, and have as many points of origin as possible. After introduction most of the bias seen will be a result of the majority of the information it is encycloping.
You must be a leftist
You ok there, buddy?
Leftism is not the same thing as communism. Thanks for answering my question lol
…and there it is. The American bias I was referring to.
…actually you’ve answered a lot of my questions. I’m all set here.
👍