Many reasonable arguments are made to counter the fear of the increasing abilities of the computer systems collectively referred to as Artificial Intelligence (AI).
Simple logic and rationales based on historical events, however, give credence to a different perspective.
First, reasonable arguments have frequently been inaccurate in the extreme in predicting the future. Aviation gives an excellent example. Initial skepticism of the utility and adoption of aircraft was initially widespread. Later, when the impact of a rapidly improving technology became apparent, predictions of an airplane in every garage were likewise radically incorrect.
With AI, there are many unknowns. Just as the potential impact of the accidental escape of a small number of African bees in South America was far more significant than anyone could have expected, something as powerful as advanced computer technology is very possibly dangerous in ways we cannot foresee.
There are other vectors that AI may follow to cause catastrophic damage to humanity. If we assume that the risk of AI to humanity is, at worst, small, then we must consider AI as a powerful tool which, like any powerful tool, is likely to be misused–or, at least, used for the benefit of a few at the expense of the many. Every technological advancement has been used by one group or nation to obtain supremacy over a competing group. To assume we will behave differently in this situation defies logic.
It is reasonable to believe that, like many technologies, the pace of technological improvement in this area will accelerate. Eventually, this will likely become exponential as each advancement has a compound effect on the next.
Danger comes not only from the fundamental changes that AI has on society (e.g., job displacement) but also in our inability to keep up with these changes. One can argue that the social fabric of society is struggling–and perhaps failing–to keep up with the changes we are already experiencing. Social media is a single aspect of the technology that is central to our lives. Few would argue that stress, anxiety, and depression clearly correlate with our adoption of this new toy.
Like the Luddites opposition to the Jaquard loom, we are unlikely to be successful in stemming the adoption of ever more capable automated machines. Many argue that we will adapt to this change as we have to all the others.
There is one difference this time. If we extrapolate the capabilities of our mechanical servants, there will come a time when they can take over all of our physical needs. At that point, we only need do the things that we want to do.
The question then will be. Is that a good thing?