Відео

straight and informative post

Very intresting and explained very well, thanks

Good video. 🙂 I think sound quality is the #1 thing you can improve. I can tell it's a bit better from the first video, already!

FLOPs = Floating Point Operations A "floating point" number is a number that can have a fractional component (like 8.563, but in binary). Computing speed is measured in the number of floating point operations performed per second (FLOPS, or less ambiguously, FLOP/s). You can count the total number of FLOPs to measure how much computation went into some batch of work, like training an AI model. The idea is that there is some unknown minimum threshold of computation that is needed in order to produce a potentially very dangerous AI model, so that threshold is used alongside the monetary threshold as a way of restricting the bill to apply only to highly capable systems created by giant corporations. If being on the hook for a giant catastrophe caused by their model makes them hesitant to release... good!!! That's the point! They should be highly confident that their system cannot cause a catastrophe, or be used to cause a catastrophe with slight modifications. These companies saying innovation will be harmed by this bill are tacitly admitting that they intend to create extraordinarily dangerous artifacts.

Great video! Well explained and nicely edited

Good efforts.

Good job ! you really have an innovative approach!

Thanks for sharing this talk, very interesting and needs to be more widely heard.