We really need a physically isolated data center for these frontier training runs.
Like a data center inside a mountain, where based on the laws of physics, no electrical signal can reach through the granite. Then also put them inside a faraday cage. And power them with a generator, such that there are no cables coming out of this data center connecting it to the external world.
OpenAI says this is not feasible due to the need for monitoring - well the monitoring server can be a separate physical machine inside this physically separated data center.
Software is soft. Use physics to contain these models.
Like a data center inside a mountain, where based on the laws of physics, no electrical signal can reach through the granite. Then also put them inside a faraday cage. And power them with a generator, such that there are no cables coming out of this data center connecting it to the external world.
OpenAI says this is not feasible due to the need for monitoring - well the monitoring server can be a separate physical machine inside this physically separated data center.
Software is soft. Use physics to contain these models.