I recently watched this small video about another amazing Google product called “Google Deep mind” .(It’s the famous program Alpha Go Ai that became famous after beating world champion in a game called “Go”)It is amazing that we can see a general AI within this decade. The main difference between Alpha Go and it predecessors was Alpha Go used reinforcement learning and neural network to mimic learning process of a human. In other words Alpha go just didn’t calculate all possibilities selected the best moves (which is impossible as the game Go has far more number of moves than the number of atoms in universe) but rather it learned by its own mistakes and got better by practicing and by watching other human players play the game. Sounds like science fiction but according to the creator Deep mind the way Alpha Go learnt to play Go game is pretty similar to how we (humans) learn anything new to us. People are worried that this might be a warning about AI being smarter than Humans but I think it is inevitable that robots will be the next phase in evolution one day and there is no need to be worried. It has happened previously and it will happen once again. As an engineer and a tech enthusiast I believe that machines (tools) created the man we know today. (imagine you have no phone, no PC or Laptop or any electricity, car or anything that human created using science and technology, you will be more helpless than a caveman who had at least few stone weapons.) It helped mankind to go beyond it’s physical and mental abilities and do amazing things. Therefore having machines that have similar abilities as a human might increase our chance of survival as a spices. but as we all know science and technology is like a coin. It has both positive and negative effects on humans. It’s up to future to decide what will be the relationship would looks like between humans and AI. You can watch the video about Alpha GO and a public speech by the creator of Deep Mind using following links.
Public lecture by Demis Hassabis
sudaspace 