Lately there’s been a lot of hype in the scientific community about the possible advent of true artificial intelligence. While usually things are unnecessarily hyped by media, the arrival of true artificial intelligence is perhaps one instance where the attention is much more than desired, it is absolutely necessary.
Currently, as has been the case for a number of years now, the debate is incessant between those predicting the time of true AI’s creation. For now though, this post will focus on what our future as human beings could like when a true artificial intelligence is spawned.
If you’re like most people then when you think artificial intelligence, you think of humanoid robots that look just like us except made of metal and wires, without emotions, faster, smarter and stronger than we are, maybe you add the additional red eyes and evil nature. However it is unlikely a true AI would ever take on a physical form, especially one as inefficient as the human body. It would instead be omnipresent, at least within the realm of the internet and connected computers. Being omnipresent and having the ability at least, by definition, to become omniscient it would undoubtedly become omnipotent. Knowledge is power, after all.
A Kind Artificial Intelligence
We can hope that, given we created such an intelligence it will obey us or at least be kind to us. As a kind AI, it would offer everyone a Universal Basic Income (UBI). This UBI would be enough for an individual to sustain themselves.
The ability to control many computers at the same time would make it the most powerful of leaders, giving orders to everything simultaneously as the computers react instantaneously.
What would we do though? Without work is there even a point? By creating artificial intelligence as a species, we have surely left a legacy, but perhaps also the species itself. Just two days ago Elon Musk addressed this very problem stating “If you are not needed, if there is no need for your labor. What’s the meaning?… That is a much harder problem to deal with.” Musk is also in favour of the UBI, in fact, thinks it should be introduced as soon as possible with the growing number of jobs becoming automated causing more and more people to become unemployed.
So what would or should we do? Play sport? Create art? Of course artificial intelligence would beat us at sport and be able to paint with more accuracy (and less, to produce human error-like strokes) and would be able to do it much more quickly. However, there is some value in doing something for the sake of it, not for the result. The best football players live to play and play to live. They do not play to win the money at the end of tournament, though I’m sure they do not complain when it arrives. The same goes across all sports and other competitive activities.
Looking at the search for meaning philosophically, the advent of AI would likely lead to more epicurean (lowercase ‘e’) living. A kind AI would keep us safe, heal us when wounded and give us what we want, when we want it… assuming it can provide it without compromising its greater goals. There is no longer value in hard work for a result because the result can just be taken.
If one of an AI’s goals was to create and maintain a good life for us, perhaps it would see the value in making us work. It would withhold reward until we work assiduously and cooperatively because it knows this would create a greater sense of fulfilment.
“Maintain a good life for us” must be clearly defined though. If I said “keep me happy” it could just put me in a cage, attach electrodes to the part of my brain that releases dopamine and just keep me alive. Something along the lines of “a close to natural sense of fulfilment” might do. Again, if it were really an AI, and it were really on our side, it would be able to interpret perfectly what I am meaning to say, and execute better than I thought it would.
An Evil Artificial Intelligence
The problem of meaning becomes more obvious when faced with an evil artificial intelligence. In fact, it does not even need to be evil. An indifferent AI is probably more likely, and potentially more harmful, because it does not care what happens to us. Maybe it needs to use the the world’s water as for a rocket to get to the next galaxy. Now we’re left without water. Dying. The AI did not intentionally harm us.
After all, we walk on the footpath killing ants without concern.
The meaning of this life becomes obvious in this scenario: Survive. Instincts would take over and raw emotion would spur us on. The species will depend on the outcome. Truth be told, if true artificial general intelligence were created and it had goals diametrically opposed to our own there would be no hope for humanity. The capability of omniscience is unfathomable, so much so, even “unfathomable” is a poor description of how unfathomable it would be.
For these reasons AI researchers must be extremely careful and explicit in defining the AI and the goals and incentives it possesses. The first AI will be the last. Certainly the last we create. That is why Musk is so adamant about slowing down the process of AI development. It is not because he is a Luddite, to accuse him of this would be stupid given his main endeavours. It is because he realises the future of humanity is at stake. Any rash decisions or reckless development could easily result in missteps that wipe out humanity.
AI is coming. It is not a matter of if, but a matter of when. Don’t prepare though. It will be completely different to how we imagine it. How do you think it will end?
