… By setting up the Fitbit Air, you’re agreeing to: Fitbit Terms of Service Google Health Additional Terms of Service Google Health Data and Privacy Policy formerly Fitbit If you’re concerned about your health data being used to train Google’s AI, that option is turned off by default and must be opte… …
… We’re not taking a detailed bill of materials. The supply chain connections themselves are revealed from the data as we connect this supply chain graph. Then, the other bucket of rights are around learning with pre-trained models. What we can’t do is memorize our customer’s data. …
… I don’t think you’re going to be done with training, by the way. I think the inferencing stuff is going to be additive. Do you need all the new data center build-outs if it’s inferencing instead of the training? …
… There’s a world in which the website is just the backend database for a very complicated Google system that is doing all of the things that you’re talking about. Is there a hedge against that or is that train just on the tracks? It’s a little bit of a hedge against that. …
… To me, the excitement about space-based data centers suggests typical Silicon Valley groupthink that gave Musk a chance to execute a hail-mary pass. Debt bets By stapling several failing businesses onto SpaceX, Musk is attempting to become too big to fail. …
… This is the first major acquisition that Netflix has done. The biggest one before this was for the Roald Dahl library for more than half a billion. That’s up against the Paramount side, where the cool has been gone for years, or for decades even. …
… If Anthropic and OpenAI and the rest aren’t engaged with you, how does this work? It’s a really great question because one of the concerns or questions that I actually have about AI comes back to that veracity of the training data. …
… The social media companies all said, “Well, you can use it for good or evil. We’re neutral platforms.” It turns out maybe they should have thought of some of those harms earlier. Look at the AI companies today. Who knows if training is copyrighted. …
… The real innovation in my opinion is, first, they’ve gotten better at pre-training and better data inputs, but the second thing is that we’ve basically taught these large language models to do better agent loops. …
… It can see all the data. Yeah. So, you can just describe a generalized outcome. You can retrain based on that data and you have perfect visibility into the actions, reactions. The retraining output you don’t have perfect visibility into, but you can iterate around that. …