Artificial Intelligence is reaching the exponential curve point at an accelerating space and before we know it, we could be faced with an “Ex Machina” era. To date, the systems have been designed to effectively learn and react to the inputs that are provided them, using algorithm’s to determine the best response (each time). The inputs are then optimized to bet better and therefore the responses get better. Basically they are as good as what you feed them.
Many firms today are looking to actually leap frog current generation platforms and design brain like thinking capabilities that truly allow the AI to get smarter based on previous memory and exercises…learning. Alphabet (Google’s Parent Company) has been investing in their DeepMind AI and believes they have made the first meaningful breakthrough. They have created the ability for the AI to remember past knowledge…meaning its remembers and learn on top of that memory. So here’s how it played out. They took the AI and allowed it to learn how to play a 1980’s Atari video game, they then switch the game and the AI was smart enough to learn from Game 1 and apply leanings to Game 2.
More to come here…but hold onto your seats and your brains…you will need them to have discussions with machines soon enough.