It's not convincing to me yet. Not that I know more, but I don't understand and haven't heard of a good explanation. Right, some technologists (from Google mostly I think) do not fully understand some of the unexpected learning from these AIs, but I believe those are still "narrow", or specifically deep learning AIs -- meaning they go down one very specific path and not the general path. It's very specialize and I don't see how being great at one area is a real danger to human. No question, machine will be always more efficient because they are very task focused, whereas human are not. But you do see some example of what happen when human are (by accident or by brain damage) - severe autism with specialized abilities.
This is posted in hangout, but I feel it belongs here as well. This is something that will impact everyone far more than Trump or Moore, AI is now at the point it can pretty much solve vast majority of rule based games systems. Trump supporters are still dreaming about taking back jobs from China, India or Mexico, LOL. You can bet there will be AI miners within a generation that will take all the mining jobs from humans. Zlpha Zero learned chess from just rules in four hours and then destroys the strongest chess playing computer stockfish, it duplicated the success in Go and Shogi. Welcome our new google net military commander. Who needs human analysts in the future? AI AlphaGo Zero started from scratch to become best at Chess, Go and Japanese Chess within hours https://www.nextbigfuture.com/2017/...chess-go-and-japanese-chess-within-hours.html The AlphaZero program developed by Google and DeepMind took four hours of playing against itself to create chess knowledge beyond any human or other computer program. It could beat any person and beat the best World Computer Champion Stockfish 28 wins to 0 in a 100-game match. Starting from random play, and given no domain knowledge except the game rules, AlphaZero achieved within 24 hours a superhuman level of play in the games of chess and shogi (Japanese chess) as well as Go, and convincingly defeated a world-champion program in each case. Self-play games are generated by using the latest parameters for this neural network, omitting the evaluation step and the selection of best player. AlphaGo Zero tuned the hyper-parameter of its search by Bayesian optimization. In AlphaZero they reuse the same hyper-parameters for all games without game-specific tuning. The sole exception is the noise that is added to the prior policy to ensure exploration; this is scaled in proportion to the typical number of legal moves for that game type. Like AlphaGo Zero, the board state is encoded by spatial planes based only on the basic rules for each game. The actions are encoded by either spatial planes or a flat vector, again based only on the basic rules for each game. They applied the AlphaZero algorithm to chess, shogi, and also Go. Unless otherwise specified, the same algorithm settings, network architecture, and hyper-parameters were used for all three games. They trained a separate instance of AlphaZero for each game. Training proceeded for 700,000 steps (mini-batches of size 4,096) starting from randomly initialized parameters, using 5,000 first-generation TPUs to generate self-play games and 64 second-generation TPUs to train the neural networks. In chess, AlphaZero outperformed Stockfish after just 4 hours (300k steps); in shogi, AlphaZero outperformed Elmo after less than 2 hours (110k steps); and in Go, AlphaZero outperformed AlphaGo Lee (29) after 8 hours (165k steps). The game of chess represented the pinnacle of AI research over several decades. State-ofthe-art programs are based on powerful engines that search many millions of positions, leveraging handcrafted domain expertise and sophisticated domain adaptations. AlphaZero is a generic reinforcement learning algorithm – originally devised for the game of Go – that achieved superior results within a few hours, searching a thousand times fewer positions, given no domain Share this: Facebook88WhatsAppRedditPinterest1
I've known people in the past that have played online against others. Is that all but dead at this point? I would not want to put real money up against a potential program.
Companies like Google and Apple are working on algorithms that can analyze expressions. The hardware behind depth mapping of the face is already down pat. It's scary to think that something like a iPhone X or Microsoft Kinect already has all the hardware technology to read facial expressions the way you described. It's just the computer scientists are now working on the software algorithms to read those maps.
It really is pretty creepy when I walk in my son's room while he is playing his XBox One and "Hello Daddy!" pops up on the screen. I don't want my son's XBox One to be able to recognize me. I'll be its first target once the Singularity hits.
