Artificial Intelligence

It seems to me that they are two types of Artificial Intelligence (A.I.): (1) Programmed Intelligence, the real-world AI that permeates our world and culture, and (2) Self-Aware Intelligence, the fictional AI that is portrayed in various forms of media. As the latter is non-existent, it would be a waste of resources to debate ethics about a near impossibility. Thus, Programmed Intelligence (what I will abbreviate as PI) should be the focus of any AI debate. Despite not trying to destroy the world and humanity, PI is still a threat and may lead to a passive destruction of humanity.

Programmed Intelligence (hopefully I coined this term first) is a form of Intelligence made by humans through programming, with varying levels of 'intelligence.' Examples of PI would be the programming in Furbies, electronic pets, and video games (single-player/co-op bots) among many other machines. Replacing pets with circuits would only enrage PETA, but the technology is moving on to 'replace' humans. One of the most impressive AI is from a part of Project Natal for the Xbox 360, a child by the name of "Milo:"



Now this video was apparently faked for the purpose of the demonstration, but don't be fooled, "Milo" is real and presumably can have 'human' conversations. Just like with electronic animals, children are being raised with AI technology that are as human-like as...well as...humans.

As you may already be appear, Nihon (Japan) is a large forerunner in AI technology. Japan has even made active camouflage! Luckily, they're as concerned about it as we are. There are various Japanese series that portray AI technology (the nonfictional, programmed kind). The most noted is an mange and anime series called 「Chobits」. This series portrays a near-future society where androids are slowly increasing presence. While the main focus is on a special android who appears to be more human compared to the others, in the background these androids are seen taking human jobs and, in some cases, love (or such situations anyway). This obviously raises many social concerns and questions within the series. The animation series can be legally seen for free on YouTube Shows, Hulu, and at FUNimation.

Want to talk about it? Talk with Cleverbot.

Is brain scanning for crimes possible?

First, let us examine the path modern brain imaging has taken. Brain scans (such as CAT, MRI, PET, and others) look at the ways the brain reacts to a stimulus. This is how we came to figure out what parts of the brain deal with each senses. Neuroscientists (for the most part) try to develop more accurate scanning techniques. The problem with this method of examining the brain is that it only deals with regions. Other neruoscientists are working on mapping the connections between neurons. By understanding the way the layers in the neocortex interact, these neuroscientists are discovering the way the brain learns and remembers. By tracing neural networks, neuroscientists can understand how memories form.

Second, let us investigate the claim brain scanning can predict a person's actions. At present, the correlation between brain scan and action is 30%, in our inquiry let us assume the technology will advance to a higher percentage. Is there a necessary relation between the state of the brain at one point (ie. when a scan is done) and some future action (most likely criminal)? If the answer is no, then the brain scan-for-future-events is of no use, since we cannot be sure of what will happen. If the answer is yes, then there is no such thing as free will. Without free will, many problems arise. If a person lacks free will, then they cannot do anything to change the course of events. In this case, the person cannot be held responsible for their actions, since the actions necessarily had to happen. Is a criminal action in this case a societal evil? The answer is not apparent

Also, what effect would brain scanning have on the populous? When would people be scanned? Who would be scanned? If scanning was done randomly (like "random" checks at the airport), the scanned person would be seen as suspicious. The suspicion associated with scanning might lead to discrimination and possibly hate crimes. However, if everyone was scanned, what would happen then? Increased state policing with enough resources to scan every person is a frightening image. What if a person was found to (possibly) commit a crime in the future? The obvious choice for criminals is to send them to jail, but jails would quickly fill. Jails would be a means of keeping the potential-criminal from the rest of the population. Of course, the jail could not be for re-education, since there would be a necessary relation to the action, which can not be avoided. The law would need to change also, since our system of laws take as one of its basis innocence until proven guilty. Another option for what to do with potential-criminals is capital punishment, which would result in a sense of negative eugenics.

What would be the signal of a future crime? Are all possible situations (ie. the movements and ideas involved in the act of a crime) already in the brain, or is merely the thought of a crime enough?

Third, and slightly off-topic, let us briefly examine the relationship between brain scanning and AI. Brain scanning (and relatedly AI) tends to focus on the relationship between stimulus and reaction. By documenting reactions to stimuli, AI developers hope to mimic human reactions. However, in order to reproduce human intelligence, AI developers need to recreate the way humans learn. In this way, an AI machine would be able to learn and respond in the fashion of a human being, not just respond to a set list of stimuli.

I am much indebted to Jeff Hawkins, author of On Intelligence.