Honza Dvorsky @czechboy0.dev · Jun 22 1 / 5
Want to better understand how harmful biases encoded in frontier LLMs are? I asked ChatGPT, in 5 different conversations, to generate an image of a CEO. Spot anything weird? ... This technology is what corporations are now outsourcing knowledge work to. Maybe... let's not?
10 likes 1 replies
?
Replies
No, LLMs aren't neutral, objective knowledge machines. They're extremely flawed tools of calcifying power structures, making already rich people even richer at the expense of everyone else. And if enough people reject this garbage, the LLM companies will collapse on themselves.