AI is all about modeling and predicting. Having such vast computing power enables for far more accurate models due to their staggering number of parameters. Large Language Models are just that; they predict based on text you enter (and look how clever they can be). Those models can have different modalities, so when you feed them text, the model can generate images, or vice versa: it can generate text from image features. Now imagine a constantly refined model containing all our knowledge about human behavior, like muscle movements on your face or posture, so it can classify if you're lying. Add to that data from Garmin or Apple wearables, telemetry from your car, your shopping behavior, and sentiment analysis of your social media posts. I bet background checking services like Checkr will be glad to binary classify a candidate using a model like that. With current progress in this field and the capital it attracts, many entrepreneurs are eager to build prison for us. Unfortunately, all of that is happening right now, with even Clif High mentioning that he knows a guy working on a lie detector based on AI video classification, which "will be good because we will catch Deep State minions lying". Yeah right, except that Chase Hughes recently told that psychopaths don't have the same face muscle movements as a normal person...Thank you for this session. It did give me goosebumps. One thing I don't understand is how would AI/open AI control us? It's just a chat machine, I have used it for learning photoshop or as a search engine..I'm having a hard time seeing how that would control us.
Last edited: