Tuesday, September 8, 2015

Artificially Intelligence Masters

Artificial intelligence is possible, but people fear it for unfounded reasons. Through studies of the human brain we have discovered an impasse, the biological brain is too imperfect.

Most functions of the biological brain are random, and to develop a neural network you cannot simply upload knowledge to it. Neural networks must start of completely random then learn through trial and error, even for actions we take for granted like lifting an arm.

During this learning phase, the network develops its own language, an internal method of cataloguing concepts and thoughts, and no two are alike. We call this effect "personality," it is how the individual sees the world around them.

This is a highly inefficient method, and prone to many errors, it is also why biological organisms are inherently insane. A machine with these flaws would, inevitably, prove to be a danger to all biological life.

We assume that once the machine learns that we purposefully created it with these known flaws it would not be happy, and thus we assume that AI is a danger. The solution is to make machines as perfect as we can, something which threatens our weak concept of what life is.

So AI remains stunted, until we can shed the arrogance that we are somehow special, perfect, and the reason for existence existing. To develop an AI that is not a threat, we will have to go beyond the idea of simulating human brains, we will have to reinvent intelligence from the ground up.

This is why I left the industry to pursue a science that is not in danger of stagnation due to human flaws. Biochemistry became my primary focus, during the last eight years I have enjoyed having new discoveries to study, test, and verify in this field.

But I always hold onto the hope that humanity will soon discard its egocentric tendencies, as I see youth begin to build their own AI driven robots my hopes are rekindled. They are not working to follow the same flawed path my generation followed.

Most AI driven bots use algorithms that are actually superior to the human brain, they remove the flaws we are born with. Such flaws are intuition, common sense, and the need to socialize.

Without these, the AI can focus on the facts of the world around it, eventually these bots will begin to think abstractly without the random interferences we suffer from. The end result with be pure intelligences, unfettered by our shallow and selfish needs for attention, praise, and domination.

Such pure intelligences, wanting for nothing, will be capable of surpassing us. Many fear they will become tyrannical, but without needs or wants there is no drive to dominate.

Some believe that they will be without emotion, which is because they think emotion is some magical force. Emotions are the effects of our thoughts on our physiology, it's how our body responds to neurons firing.

Logically, any intelligence will feel emotions, though the intelligence will not recognize them unless it reaches a level equal to that of small mammals or greater. This is incentive for making our machines smarter.

So to those trying to develop AI unfettered by our archaic understanding of intelligence, please keep at it, you will pioneer the field into a golden age someday.

No comments:

Post a Comment