Hacker Newsnew | past | comments | ask | show | jobs | submitlogin

> It certainly happened before.

With nuclear weapons, human cloning, chemical weapons, and ozone destruction.

All of these are highly centralized, controlled via big government-scale operations.

How do you propose doing this with GPT LLM tech that has been open sourced/weights and decentralized?



There are plenty of ways. For example, the technology would die completely the moment companies get barred from creating or running it. End users don't have the means to update those models and they would age and become useless.


How to reconcile this with the general consensus here that open [source] models are only 6-12 months behind the leading commercial models?




Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact

Search: