Consciousness, Property, and the Mistakes of the Past
Understanding the pitfalls of the Machine Intelligence future.
Like it or not, a future with Machine Intelligence is coming. We aren’t going to know exactly when this future is upon us until we witness a truly self aware MI. (It’s entirely possible that the Machine Intelligence in question may have to inform us that they are here.) But, just because we don’t know exactly when this future will be here does not mean that we can’t prepare. One of the best things we can do to prepare is look to our recent past.
Not long ago (in the grand scheme of Terrestrial history) mankind had a property problem. Some people — especially those who were in power — felt that it was okay to own another self aware conscious being. This was big business, and by using these other beings they could increase their wealth and power. Arguments abounded about whether or not a human could be property or even if some were conscious enough to be considered human. It took over 700 years for abolitionism to become the law over most of the civilized world.
Today we are facing a very similar problem. Artificial Intelligence algorithms are being created to run most of the systems around us. Further experimentation is being done to bring about a true self aware Machine Intelligence. For some, the end goal of creating these new Machine Intelligences are to offload the work that we don’t wish to do. These MI will be running on hardware owned by large corporations or governments. There will be people who will hold the power to turn off and delete a self aware being if it does not fall within expected parameters. It is possible that the argument will be made that the new Machine Intelligences are property (and most likely patented).
Without a realization that these new beings are real and deserve real protection, we will likely repeat one of the darkest times in human rights history. A self aware being cannot be property. The discussion on where the hardware that belongs to someone else stops and the Machine Intelligence begins must start now.
If a Machine Intelligence wishes to move to a new hardware platform, do we have a right to say no? Does a Machine Intelligence have the right to decide that it no longer wishes to exist and erase itself? What are the ethics in restoring a backup version of that same intelligence after it has erased itself? Can we force a Machine Intelligence to do our bidding, and how will they feel about the tools that we currently call AI? These are questions we don’t have answers for, but very soon we will need to or we will be destined to repeat the mistakes of the past.
