- Speakers
- Scott KarsnerGuestSharonHost
The Massachusetts regulations really build on some laws that were passed in California and New York and Illinois, and they have a few different mechanisms.
The ones that emulate the previous legislation ask the AI companies to report any safety incidents within 15 days, and then they also have to have kind of a safety policy or safety protocol.
And so where Massachusetts wants to go a step further is not just looking at the safety protocols, but what's happening inside the software itself.
And so they introduced the idea of an independent evaluator that would really be able to open up the hood and poke around the inner workings of these AI models.
They'd be able to do that every four months, and what they're looking for is something built into the model that is not are they biased against a particular job applicant or might they cause delays of flights at Logan Airport? They're looking for large scale potential problems, like a billion dollars or more in financial consequences or 50-plus people that are killed or injured because of something that the AI does.
The Massachusetts regulations really build on some laws that were passed in California and New York and Illinois, and they have a few different mechanisms.
The ones that emulate the previous legislation ask the AI companies to report any safety incidents within 15 days, and then they also have to have kind of a safety policy or safety protocol.
And so where Massachusetts wants to go a step further is not just looking at the safety protocols, but what's happening inside the software itself.
And so they introduced the idea of an independent evaluator that would really be able to open up the hood and poke around the inner workings of these AI models.
They'd be able to do that every four months, and what they're looking for is something built into the model that is not are they biased against a particular job applicant or might they cause delays of flights at Logan Airport? They're looking for large scale potential problems, like a billion dollars or more in financial consequences or 50-plus people that are killed or injured because of something that the AI does.
The rest of this transcript — segmented and speaker-labeled, so you land on the exact moment something was said
Search every transcript — by keyword, by phrase, or by meaning, across every show Radar indexes
Trends — what is surging across podcasts, measured against its own baseline
Alerts — when a name you follow appears in a newly indexed episode
No account is needed to search Radar.
