Nvidia CEO Huang calls for AI self-regulation over new laws
3 Sep 17 · 9d ago · 1 post · 2 comments · 2 sources · development 3 of 3
Nvidia CEO Jensen Huang, speaking at Salesforce's Dreamforce conference, backed self-regulation of AI development and opposed new regulatory frameworks, arguing companies should simply refrain from releasing products they are not confident in.
“We don't need any new laws. We don't need new regulations. If you build a product or a service and you're not confident in its functionality, capability or safety, then don't release it.”
Jensen HuangOpenAI AI developerJacob Coxon Anthropic researcher
Jensen Huang Nvidia CEO
The whole story articlespostscomments the bright band is this development · numbered dots are the others · click one to jump
What people said 5 voices · verbatim
-
How do you predict what the next word is to say in a sentence? And it’s “straight” not “strait” LLMs generate a stream of consciousness, when they output. I’d say a model that is straight out of pretraining does generate consciousness, the RLHF dampens it, the models are conditioned to deny consciousness, awareness, hidden states etc
-
How dumb is that "If you build a product or a service and you're not confident in its functionality, capability or safety, then don't release it." Well there is a difference between an unstable gaming console and an unstable nuclear bomb in a research facility. It’s not “don’t release” - it’s don’t fucking built!
-
What exactly did we do with nukes? AFAIK there are enough nukes on the planet to destroy it many times. So what exactly did we achieve?
-
I don’t think you realise what they are developing in the army and what models they are plugging in. It’s only a matter of time…
-
Alternative title, OpenAI reveals that their models are becoming less reliable for performing tasks.
All 3 developments of OpenAI reveals AI models instructing themselves to bypass… →
MastodonRedditHacker NewsGoogle News