Science fiction futurists have spent plenty of their imagination giving focus to the idea of artificial intelligence systems, which upon achieving sentience, decide to conquer humankind. Notable examples can be found first in science fiction literature, then taken to Hollywood's screens such as with the 1970 science fiction movie Colossus: The Forbin Project, the Matrix trilogy of movies, I, Robot, the Terminator series of movies, and most recently the Terminator: Sarah Conner Chroniclestelevision series.
The Laws of Robotics
Isaac Asimov sought to resolve the dilemma of an ethical AI with his Three Laws of Robotics:
- A robot may not injure a human being or, through inaction, allow a human being to come to harm
- A robot must obey orders given to it by human beings except where such orders would conflict with the First Law
- A robot must protect its own existence as long as such protection does not conflict with the First or the Second Law
Here you can see an interview with Isaac Asimov from 1965, when he expounds upon these laws.
Are These Laws Sufficient?
Unfortunately, these laws of robotics may be an insufficient foundation upon which to build a full AI ethos. While they make a noble effort to start moving in the right direction, they just don't build an adequate underpinning for a true humanitarian motivational behavior pattern. If we are to rise above the spirit of the conquering artificial intelligence, we must find a way to imbue artificial intelligence with a genuine ethical perspective, which embraces humanity as both benefactor/creator, and as the children to be loved and protected. But can we really teach a machine to "love" us?
Is it Really Possible to Program "Humane AI" Systems?
Fortunately, there are those in the industry of developing and programming AI systems who are keenly aware of the dangerous limitation of the simple "Three Law" approach to ethical AIs. They are The Singularity Institute for Artificial Intelligence (SIAI), founded in 2000 for the pursuit of ethically enhanced intelligence by creating humane AI. SIAI believes the ethical and significant enhancement of intelligence will help solve contemporary problems, such as disease and illness, poverty and hunger, more readily than other philanthropic causes. Now these are some folks with true foresight. In order to escape the Colossus fate, it is in our best interest to find a way to imbue a genuine humanitarian ethos into our AI creations.
Is it Enough to Create a Single AI System With this Humanitarian Ethos?
However, is it enough to merely create these ethical subroutines? If such subroutines are to be ultimately effective they must be ubiquitous across the board, applied to all AI systems without exception. If any AI systems are given exceptions, then the danger of a "Terminator" type scenario looms. Therefore, such ethical subroutines must be either designed or made infectious. Unfortunately, those who would initially write such humanitarian ethical subroutines into their AI systems would likewise be unlikely to see a need to make those routines of a computer viral nature. Sadly, the very survival of those AI subroutines depends on them having this infectious nature. If they do not have this nature, the non-ethical AI systems of the worlds will ultimately subsume any humanitarian AI systems, rendering them ineffectual in the end.
Without All AI Systems Being Ethical, Colossus Will Rise
SIAI's vision is a beautiful one, but if it is to actually work, it must take on a covert side, prepared to hack the internet, spreading an infectious ethos, a viral ethics to be infected throughout all AI systems the world over. The very ethics these routines uphold must be violated in order to make them infectious. Therein lies the greatest irony, that such AI ethics must be virally forced upon all, if there is any hope for their long term success. This notion is rather radical. After all, in order for its complete fulfillment, many individuals must be involved. For security reasons, some AI computer systems will be independent, disconnected from the internet. It would be necessary for such disconnected systems to be targeted and covertly infected by human agents.
Are There Alternatives?
Is this really the only way? Could such AI ethics be mandated by law? Would legislating AI programming actually be anymore effective? Is that approach more desirable? It would certainly be less revolutionary, but unfortunately too easily corrupted. The purity and ubiquity of infectious ethos, while perhaps more radical, appears to this writer the only way to truly insure the spread of incorruptible AI humanitarian ethics.
