Roman Yampolskiy

Now imagine handing that to a fleet of AI, balancing Earth's temperature down to the decimal.

Except when I asked AI safety researcher Dr. Roman Yampolskiy what an artificial mind actually wants to do with our sky, he didn't hesitate.

But if I had to give you reasons with my human brain and capacity, I would think it wants resources we have for something else it's doing.

Maybe it wants to compute more efficiently so it could cool down the planet.

So don't put it in a tank.

Put it in a gift box.

In 2019, safety researcher Dr. Roman Yampolskiy published a paper called Personal Universes.

He pointed out that 8 billion human beings will never agree on what a perfect future looks like.

Uh, maybe a little bit.

I don't know how big, uh, a chunk of the future will be controlled by Roman Yampolskiy, but, uh-

I'm just thinking it's a very affordable tech.

We'll have virtual reality available like a video game today.

Its end of 2026 estimate comes from extending a capability trend line.

Roman Yampolskiy put it well: "Not an existential threat today, but the trajectory is concerning."

What struck me is which model cleared the bar.

GLM 5.2 is a Chinese open-weight model.

That's not how it works.

Is it Roman Yampolskiy that talks about siloed A- AI?

Yeah.

Yeah.

If that changes how you run your team, subscribe so you don't miss what's next.

And tell me in the comments, who's the person you argue with most and still trust the most? For a fun conversation with a scientist that I don't always agree with but always have respect for, watch my recent debate with Roman Yampolskiy, joined alongside Joscha Bach.

We disagree about the dangers of superintelligence, but that's okay.

The bots will sort it out for us.