A researcher creates a new kind of AI that starts to demonstrate complex, intelligent behaviors. No, we're not talking about Skynet. A researcher with ties to MIT and Harvard has actually created what he says is a powerful artificial intelligence that, in simulations, was able to "walk upright, use tools, cooperate, play games, make useful social introductions, globally deploy a fleet, and even earn money trading stocks, all without being told to do so," as the video below explains. And here's where things get even weirder: the AI, called Entropica, is based on equations derived from the second law of thermodynamics, suggesting a link between entropy and intelligence. Some observers are skeptical. Is it for real? Only time, and entropy, will tell.
No comments:
Post a Comment