- You can’t have unbiased AI without unbiased training data.
- You can’t have unbiased training data without unbiased humans.
- unbiased humans don’t exist.
show your work. 1 especially seems suspect. Especially since many AIs are not trained on content like you are imagining, but rather trains itself through experimentation and adversarial networks.
Even how it trains itself can be biased based on what its instructions are.
Yes, and? If you write a bad fitness function, you get an AI that doesn’t do what you want. You’re just saying, human-written software can have bugs.
You’re just saying, human-written software can have bugs.
That’s pretty much exactly the point they’re making. Humans create the training data. Humans aren’t perfect, and therefore the AI training data cannot be perfect. The AI will always make mistakes and have biases as long as it’s being trained on human data.
deleted by creator
AI judges make a lot of sense, that way everyone is treated equally, because eveey judge thinks literally the same way. No corrupt judges, no change in political bias between judges, no more lenient or strict judges that arbitrary decide your fate. How you decide what AI model is your judge is a whole new can of worms, but it definitely has lots of upsides.
Perhaps when we have real AGI, but I wouldn’t want an LLM to decide someone’s fate.