Looking at my feeds this morning , I noticed NASA is thinking about robots on the moon in 2020. That is at least a step toward doing what I know should be done, and that is to send the lathe and a way to operate it.
I was working on the nature of biogenesis and started creating some software and considering what hardware or specifically "wetware" ( self replicating circuits and memory ) that might help. As a result I was considering a specific method for self replicating memory. What I realized is that I could combine several technologies and have a living liquid brain that would be vastly more intelligent than a person. In the process of analyzing how it operated, I realized it would be self programming. This is the most intriguing aspect. I had wondered how a person could design a usable ( not requiring programming ) robot brain that was at least as bright as a squirrel and now I know.
I am absolutely certain I can make a growing liquid mind. I am now considering what it actually means before I think about making it. I had some general ideas before and they would have worked with quite a bit of my time and effort. This is virtually a breeze in comparison. I could build the first elements for almost nothing and expand it with fairly common raw materials. It has one drawback, and that is speed. It would function below biological speed, but function at complexities that are only limited by resources. I have thought of some things that would speed it up, and ways to give it a base program to start.
By analyzing the properties of the device in simulation, I found it would do very similar things to other biological intelligence. It is the complexity that makes the break over point between the brain of a fly and a dolphin. I am very psyched by this discovery as it certainly functions as a model for a distributed emergent intellect.
I also discovered a very odd relationship in the patterns of cell growth and division. I also read some background on http://en.wikipedia.org/wiki/Bacteroidetes, which is very unusual as they have such a small size that they can pass through bio filters, which is a little scary.
I am also working with one of my latest compulsions, which is revisiting all of my perl code, perl cgi, RegEx ( which seems to be the 'cool' shorthand to say Regular Expressions ), and revisiting FSA or finite state machines. I am also doing the programming and design for my little ant game as I wait for things to settle down in the focus of antfarmgl.
ADDED: This may take some getting used to. The idea of a mind that is 100's of times more intelligent than a person. It is certainly coming and many of the new super computers can be used to do things that are astounding. I don't think winning any game is beyond them. It is the idea of AGI ( artificial general intelligence ) that eludes this type of design. The ideas I have will make an AGI machine of incredible potential. I can't say whether it makes me sad or happy, it is a mixed emotion to see such a fantastic possibility and such a degradation of the position of human intellect as a measure of humanity. I have heard others say that AGI is only 30 years away and I suggest it is about -4 hours from being designed, in other words, it has already been designed and will likely be completely implemented in less than a few months.
0 comments:
Post a Comment