Racism and AI guardrails.

in ai •  last year 

image.png

Dudes now taking their bizarre N-word hypotheticals to chatbots. Haven't they suffered enough already as fake girlfriends for Incels?

image.png

New AI doomerism: "this chatbot won't say the N-word, what have we done? What have we done? We didn't listen"

I don't even mind the premise in general, but please stop like ChatGPT having some guardrails around racism is one step away from doomsday or something. It is just a language model. We aren't putting it in charge of the nuclear stockpile. It isn't AGI. It isn't sentient. It is a language model that without such guardrails can easily start generating really racist stuff.

If you want that fulfillment in your life, looks like Elon has you covered...

Authors get paid when people like you upvote their post.
If you enjoyed what you read here, create your account today and start earning FREE STEEM!