As AI models get smarter, the harder it becomes to correct them. Their values become ingrained. 

Imo, this is more negative than positive because it will solidify the models biases and undermine free speech.

Thoughts on paragraph.xyz/@antidote

Agents and Intents @NEAR | prev investments and BD 
@ZkSync, New Enterprise Associates and Balderton Capital

how can a model have values? or do you think it’s just mimicking them based on pre training?

Yes, training data will lead to bias and that bias leads to political / societal values.

Do you think it’s possible to have bias-minimized training data when all humans produce biased things ?

I am sure it would be theoretically possible but it would probably resut in a model that is worse than biased models simply because you have to more carefully curate and restrict training data