For the season finale of The Mirage Project, we travel to a planet that has been turned into one giant machine... 'The Last Machine' - A place where all biological life has been wiped out...
If we don't pay attention, this is exactly what might happen to us... We have been warned...
"The development of full artificial intelligence could spell the end of the human race."
Stephen Hawking
"AI is our biggest existential threat."
Elon Musk
Here's the problem: We are currently working hard to improve the intelligence of our machines. For good reasons: Better, smarter machines will help us to address huge challenges. How can we feed the ever-growing population? How can we develop new cures for illnesses? How can we move on to new, safe sources of energy?
Today, we are using intelligent machines - computers - to build the next generation of those machines. Those machines will become smarter and smarter. The time until the next generation of machines gets shorter and shorter. And one day we will build the last machine ourselves. This machine will be able to create smarter and smarter versions of itself.
And pretty soon we will be faced with an Artificial Intelligence that will look at us the way we look at ants. Not with pity or hatred. But with complete disregard. What an ant wants is of very little concern to us. We let them do their thing. But if they get in our way we exterminate them without giving it too much thought.
The moment all goes wrong
We all know this scene below - but it's just too good not to show it again in this context...
Terrifying - and likely to occur?
Here's a great TED talk with neuroscientist and philosopher Sam Harris about this threat. In his brilliant presentation, Sam says that a scenario in which we might lose control over general Artificial Intelligence is "both terrifying and likely to occur". The question is, can we build AI without losing control over it?
The King Midas problem
Is there anything we can do about this threat?
Let's assume that we build these machines with our best intentions. We want them to work to improve the lives of all humans. Now in theory, that sounds like a great idea. But if we are not really careful things could go terribly wrong.
Just think of King Midas. He asked the gods to turn everything that he touched into gold. The gods listened and gave Midas this skill. And it worked: Midas touched his beloved daughter - and she turned into gold, cold and dead. Midas touched his food - it turned into gold. And so the richest man of his time-starved to death.
Midas' mistake was that he did not really think through what he was wishing for. He got exactly what he wanted. And it broke his heart and ultimately killed him.
Initially, we might still be able to interact and understand General Artificial Intelligence. But after just a few generations it will become so much more powerful than our limited human minds, and with this, so strange, that we will most likely not be able to fully comprehend what is going on in such a system.
Take the most intelligent chimpanzee and give him the best training. He will still not be able to understand how our technological society works. And if you look at the living conditions of most human apes, you see that it's just not a good thing to share your planet with a being that is intellectually superior...
There's more to come...
I hope this gave you some food for thought. Artificial Intelligence has fascinated and inspired humans for a long time. So, in our next newsletter we will share some more videos, podcasts, books and movies about AI. Until then - we hope that your toaster will not murder you! ;-)
'The Last Machine' by Sonarpilot is out now
Experience the full video in HD here
The soundtrack to The Last Machine is also out now and available to buy/stream from all major digital music platforms
Till next time, take care!
The Mirage Project