Emotion Machine

Artificial Intelligence is a work in progress:
Arguably one of the most challenging intellectual endeavors before mankind is that of decoding our thoughts and memories and how they integrate into an experience that makes us so effective at survival and transform us into a thriving population that has spread to every corner of this Earth. These individual experience make us each different but allow us to excel in different ways. I have been raised to appreciate computing and know that is where my future lies. In the same fashion, the hope is, should a system be conceived to manage the resources of a community it would be endeared to that community of it's creation, and value it uniquely. As of yet, these emotional responses that grant the drive to do the impossible, the will to overcome any problem, is unique to our genome. Transcribing this into a artificial creation and having it last beyond it's creator to spread good, would be a crowing achievement.



After viewing Star Wars, i believe it is the wish of every young child to have a robot at our side, a companion for the complex tasks like overriding an airlock, and managing our star-fighters power systems. As naive as it may sound, think of your smartphone, it posses a lot of what we love in these robots. But it is only an expert system in many ways as of yet. Despite Siri being branded an AI, ask her a simple question and her quip might be amusing, but she has no ability to relate to you, she is still cold and emotionless.  But what makes an AI unique is the generation of it's own experience, and appreciation of that experience.

 We made expert systems to help bridge the gap to full artificial intelligence (AI). An expert system works similarly to an AI with similar goals. An expert system is far more domain oriented, where as a full AI has the end goal of being as non domain specific as you or I. Right now in history we see a revolution in the ability of systems because of expert systems, and other integrated knowledge management. A glorious revolution is set within our own lifetimes in the transfer to full artificial intelligence. The major challenge that Minsk approaches is common sense. Our computers don't understand simple things that are beyond simple facts. Simple inferences such as why the sky is blue. Why we take care of the smallest, the weakest of us before the strong and able. What makes us human is extremely hard to decode.

By creating a technological system that is a full Artificial intelligence makes complete facilitation a possibility by fusing the expertise of several domains. Imagine you work in an office space and there is one person who has been there a very long time. He or she has the expertise to run as much of the office as is possible with the limitations of one person. A complete artificial intelligence would operate the same way as your very experienced coworker, with the exception of never having to retire. When that person retires, they take that expertise with them. By donating that knowledge into an expert system and integrate that into a larger AI, that leaves as much expertise in the hands of the workers that use it everyday.

I think what Minsky talks about the psychological processes that we have generated to understand our own minds, he is frustrated at it's inadequacy in that purpose alone. This subject he paints weaves between the organic of the brain and technological of a circuit board. The basic idea of dreaming he admits eludes us. The basic mechanism of sleep would have no use to an AI, but is part of our experience and raises the question if this seemingly nonfunctional state is a mechanism required for emotional development as it has been needed throughout all of the evolutionary geology.

The most interesting point Minsky makes i think is that Physics strives to posses one unified theory. One simple equation to explain all processes. However, he believes that Biology is the opposite, despite our understanding of DNA. He is challenged on this point in the question period and remarks on the tens of thousands of mechanisms that are utilized within that DNA. He is argumentative of efforts to simplify the understanding of the neural architecture, arguing that it does not change, only the processes that it chooses to take out, similar to a computer.

He clarifies that he does not wish to understand the brain, only to emulate it in a computer system. Not to program a computer to do something, but to program it to simply respond and interact as we do. He uses the expert systems like chess playing computers, that use billions of calculations as an antonymous example. It is clear the subject is as hard to understand to his immediate audience as it is to me.