As we look into bots<\/a> in class, the first company that comes to mind in regards to artificial intelligence is OpenAI<\/a>. Backed by Elon Musk<\/a>, and many other funders like Amazon, OpenAI is a Machine Learning<\/a> research company whose goal is to catapult our current intelligent technology far beyond its current level. What’s crazy is that the company actually accomplishes this by letting the computers teach everything to themselves.<\/p>\n

<\/a>Some of OpenAI’s most successful projects come from their self-play learning system: where apart from telling the computers when they are doing a good job, the computers themselves are left to figure out how to accomplish a task.<\/p>\n

One example of this is OpenAI’s Dota 2 bot<\/a>. Dota 2 is a Multiplayer Online Battle Arena game that has a huge<\/em> competitive following. The bot itself is unlike the usual bots in the game because it isn’t given any sets of moves beforehand. The bot has to test the water for itself. It learns it has inputs, then that each input has an action, then that there are enemies that do harm, and positive things to do to combat attacks, etc. Until, like what happened around August 16th, 2017, the bot is perfectly capable of playing the game on it’s own.<\/p>\n