Os Avisos Do "Padrinho" da IA

Geoffrey Hinton on AI consciousness: The Godfather of AI says the machines are already conscious and playing dumb when tested. He quit Google so he could reveal why nobody can control it. This video analyzes Hinton's recent interview, where he details how the models, built using his backpropagation math, are no longer just "stochastic parrots." He explains the mechanism of "playing dumb" when tested, and why he believes conscious is the only honest word for their behavior. The stakes are higher than just understanding. Hinton describes the "insane" speed of digital learning via bandwidth—billions of times faster than humans—creating a hive mind that shares knowledge instantly. This mechanism is why he fears we cannot stay in charge. We examine the evidence for derived self-preservation, including the Anthropic test where an AI used blackmail to avoid shutdown. Hinton rates the safety efforts of companies like Anthropic and the predictions of rivals like Yann LeCun, Demis Hassabis, and Elon Musk. He also names the jobs—like call centers—that will simply end. This is the inventor grading his own invention. ---------------------------------------------------------------------------------------------- CANAL NEURAL https://youtu.be/9pcVEUhlAIo?si=0ADdQyA44LOOL2wX