Pages

June 7, 2015

What Happens When The Machines Get A Lot Smarter Than Humans?


From WIRED:
For the first time ever a computer has managed to develop a new scientific theory using only its artificial intelligence, and with no help from human beings.

Computer scientists and biologists from Tufts University programmed the computer so that it was able to develop a theory independently when it was faced with a scientific problem. The problem they chose was one that has been puzzling biologists for 120 years. The genes of sliced-up flatworms are capable of regenerating in order to form new organisms -- this is a long-documented phenomenon, but scientists have been mystified for years over exactly what happens to the cells to make this possible.

By presenting the computer with this problem, however, it was able to reverse engineer a solution that could explain the mechanism of the process, known as planaria. The details discovered by the computer have been published in the journal PLOS Computational Biology, along with the artificial intelligence method used to develop the theory.
Actually the title to this article was a bit misleading. Programmers wrote code to have the computer run a series of models designed to reploccate the processes required for planaria. The computer could not do this on its own. But nonetheless, it was a very interesting experiment. Computers are at their best when they are utilized for their speed in running repetitive calculations.

[...]What the computer discovered was that the process requires three known molecules and two proteins that were previously unknown. This discovery, says Levin, "represents the most comprehensive model of planarian regeneration found to date".

"One of the most remarkable aspects of the project was that the model it found was not a hopelessly-tangled network that no human could actually understand, but a reasonably simple model that people can readily comprehend," he adds. All this suggests to me that artificial intelligence can help with every aspect of science, not only data mining but also inference of meaning of the data,"
Will machines get smarter than humans? Hard to say. They are not intuitive and are not capable of inspiration. They get to achieve results by hammering away at the same problem over and over until the desired result is reached. They are only as good as the program that drives them. Can a program duplicate these qualities? I doubt it, not for some time anyway.

4 comments:

Doom said...

A truly intelligent machine? It will understand that, when working with variables, there are other variables that cannot be counted... either too small of an effect for the parameters coupled with limits on computational power, or unknown/unknowable forces at work. If you could look at every possibility, add in events that are known but belie any numerical model, and can look into other areas on the periphery of what is being studied and even areas that seem even contrary but have similar pattens in some areas, you will understand that you can't know.

Some men are impressed with science, math, and all. I do not think a machine would be. They would see the massive limits, though perhaps more function through margins. It isn't intelligence, out of machines, that worries me. It is ego, curiosity unbound by even a notion of ethics or care for humanity, centricism. It would have no capacity for sympathy, probably none for empathy. Worse, these would be inherited from their makers. Programmers, engineers, and scientists, at those levels, are not known for their capacity in any areas that would make them caring, loving, forgiving, or even good at general thinking. Which was worse, Frankenstein or Dr. Frankenstein? Or, in truth, are they one in the same, just with different form factors and function sets?

I'm, honestly, not too worried at this point. Worst case scenarios, currently, would be bad but not extinction. Close, but not quite. Unless all opened and the machine became both genocidal and suicidal. Even a psychopathic AI would still need people. A lot. Too much to risk, unless it was both of those things and yet logical enough to see both to fruition.

What I fear is empty, and worse, chairs... and those who allow it.

Fredd said...

I think it has already been proven that machines are smarter than humans: former heavyweight boxing champ Leon Spinks went head to head in a Trivial Pursuit contest with a Hamilton Beech 5-speed blender: the results spoke for themselves - Blender 6, Spinks 2.

The blender has long since retired and moved to Boca Raton, FL, and Leon Spinks slowly mutated into a head of broccoli, but the Battle with the Machine lives on only in my memory...

LL said...

Machines will never comprehend the simple and most elementary of evolutionary equations, namely that the angle of the dangle is equivalent to the heat of the meat.

Kid said...

Sig, they already have robot "brains" that can learn.
I'm with Elon Muck and others. Creating a machine that can learn beyond its programming is suicide.

If we do create them, I say we test them over in the middle east.