Top Menu

The Next Frontier for Computing: Computers that can Learn

Recently, a research team made a big big splash when they publicize that they’ve produced a machine that learns from its mistakes. There is only one organisms that we know of that actually learns from its mistakes in a very conscious manner. That organism would be us. That’s right, homo-sapiens. Be that as it may, it seems that we may be edging closer to self-aware machines sooner than we think. There is a lot of benefits to building machines that can learn. The problem with factory machines and computing machines that uses a fixed set of instructions to do certain things is that, situations change, the temperature is changed, circumstances can get quite random. In other words, life is what happens when you are making other plans. Unfortunately, all of our machines are currently using systems that enable them to behave in certain ways based on certain things that we can anticipate. So in other words, you program certain parameters into the software of the most sophisticated machine and its universe is restricted to those parameters. Thankfully, when you use the actual experience of the machine as a way of broadening those parameters, you can create machines that are more flexible and can produce the desired result. In a wide variety of circumstances, this can blow away current productivity. This can truly revolutionize how factories are run and how products are made.

While this computing model is great, one of the biggest changes that is staring the technology in the face is the possibility of quantum computing. Quantum computing is that instead of using zeros and ones and linear computing, you can factor in all sorts of possible solutions and possible pathways to a solution to a problem. In other words, it’s trying to use the way the human brain works. When you move your arm for example, it may seem obvious to you that you’re just moving your arm and that you translate your thought into action. However, for that to happen there are many different neuropathways that are engaged. If you, for example, suffer a stroke, it wouldn’t be as easy as before because you were moving your arm a certain way. You were using certain pathways. However, as you probably know, when you ask physical therapist, they have ways of having the brain rewind itself. Quantum computing and neuromorphic processing uses alternate pathways for processing solutions to the same problem. So, it’s kind of like a race to whichever pathway can produce the fastest result. This is revolutionary because this can build new efficiencies and more importantly, it can adapt to a wide range of changing circumstances. This is so amazing. This technology is so amazing that its actually quite scary.