The End of Physical Experimentation

In the future, as computers become more capable in terms of processing power, they will also be more capable in terms of simulations. Experiments and hypotheses are already simulated on today's computers, removing the need to physically test them. Things such as weather patterns and earthquakes are simulated on computers. So far, the majority of these simulations have been dealing with physics or other hard sciences.

Eventually, social sciences will also gain the advantage of simulation. Simulations regarding such subjects as sociology and political science will be ran on powerful computers. Given the massive amount of variables and relative complexity of behavioral sciences, they require much more processing power and information than simple (relatively speaking) natural and environmental simulations.

Imagine a computer program in which various societies, all with their own unique ideologies, are programmed to experience certain conditions. Political scientists can design governments and simulate them on various societies to find the strengths and weaknesses. If the society responds negatively to the governmental structure, it obviously needs revising. Conceivably, this could lead to much more perfect governments, as they have throughout history relied on trial and error (or brute force) to rule their people. Eventually, governments may be prepackaged after rigorous testing in simulations.

Sociologists could study these virtual societies and individuals, while applying various stimuli and monitoring the virtual being's reactions. To most accurately represent real humans, the environment in which the virtual humans "live" in would have to be as earth-like s possible, and the daily activities and ideologies would also require much attention in order to keep the simulations as accurate as possible.

Do AI's Need Us to Teach Them?

Artificial intelligence is still a future invention, yet the foundations for its existence may be present today. Nobody truly knows how AI will be created, but some have speculated that it may be developed through social networking sites. AI's will have their own profiles, and engage in social activities like friending and messaging. These primitive AI's will evolve cognitive behavior through interaction with humans on the sites.

Basically, social networking sites will serve as virtual environments where AI's can evolve. This speculation was done by Yury Millner, a Russian web tycoon quite familiar with the internet and how it works. So maybe it should be taken with a bit of respect.

We generally picture AI's as either intelligent or not, all or nothing. But we seldom consider that they will most likely be semi-intelligent for a period of time before they truly become intelligent. It is difficult to predict whether they will need to learn through interaction with humans (as in social networking), or if they will have the ability to self-teach. I believe that they will initially need to be taught through human interaction, but the first ones will be able to teach others. In a way, the first AI's will be the Adam and Eve of all future AI's. They will teach their predecessors how to learn and thus begin a virtual evolution.

This is similar to what I speculated a few months ago, so this only enforces my prediction that social networking sites will play a large role in the development of AI's and mind clones. The only question is when.