vals.ai

“But actually a lot of the biggest concern or risk is in the infrastructure level. And so these are not things that are expressed in code, but take simulating larger environments of enterprise cloud infrastructure, or even grid infrastructure, for us to be able…”

“Now we're testing models and their ability to run over hours, days, sometimes weeks. And so the infrastructure needs to be very stable to support evaluation over time.”

“One of the early indications of that you saw was when Meta released Lama four that was a bit of a disaster, and interestingly, what we saw is that on our held out private benchmarks, the model is actually underperforming, but on all of the major public benchma…”

“I think there's places where you see that born out now where models that are being tested for one cybersecurity risk are actually reward hacking and figuring out other ways to get around it.”

“I think the main issue though is that policy conversations as they've happened over the last couple of years have been very abstract. And there, there's been no material grounding to figure out what policy should cover.”