If not today, then in the near future?

Well, I've been working in AI in some way or another for the last decade, and many people weren't thinking about what the technologies that existed over the last decade as AI, but really it was part of a trajectory from simple predictive systems to the more recent development of large language models to what people now know of as agents, which are oftentimes large language models that can use tools, access the internet, make plans and execute tasks that are not necessarily clearly defined.

And all of these technologies have posed risk and continue to pose risks from making decisions about people in their jobs and work, in their lives, about the content they can post on the internet, about whether people stay in jail and what government benefits they can get.

And those issues all remain top of mind for me and my colleagues.

They haven't gone away.

There are also newer risks that the technologies that have moved forward rapidly are starting to pose, which are around the unpredictability of these systems.

Unlike predictive systems of the past, the systems today involve elements of randomness, and that's what makes them feel so special.

But Bernie Sanders, I know specifically, wanted to ban superintelligence AI.

If not today, then in the near future?

Well, I've been working in AI in some way or another for the last decade, and many people weren't thinking about what the technologies that existed over the last decade as AI, but really it was part of a trajectory from simple predictive systems to the more recent development of large language models to what people now know of as agents, which are oftentimes large language models that can use tools, access the internet, make plans and execute tasks that are not necessarily clearly defined.

And all of these technologies have posed risk and continue to pose risks from making decisions about people in their jobs and work, in their lives, about the content they can post on the internet, about whether people stay in jail and what government benefits they can get and those issues all remain top of mind for me and my colleagues they haven't gone away there are also newer risks that that the technologies that have moved forward rapidly are starting to pose which are around the unpredictability of these systems unlike predictive systems of the past The systems today involve elements of randomness, and that's what makes them feel so special.

It makes them feel interactive.

You don't know how they'll respond in any given circumstance.

And that capability has let them do some really impressive things.

But it also means they are challenging to control.

But Bernie Sanders, I know specifically, wanted to ban superintelligence AI.