Is Elon Musk right about AI?

I've owned a Tesla Model S since 2013. It's a great car, and I will never own a car with an internal combustion engine again. I am a huge, unapologetic Elon Musk fanboy. His vision for an electrified future powered by sunbeams is spot on. He ain't no kook.

But I don't think he's right about AI.

It is generally a dumb idea to argue against Mr. Musk. He is smarter than you and me. It's like the end of that old SNL bit "bad idea jeans".

Announcer: Bad Idea Jeans.

[ shows a group of tough looking basketball players on the court ]

Guy #1: Hey, you guys ready? Let's bet these guys! A hundred bucks.. make that two hundred! Two hundred bucks!

[ fade to image on screen: BAD IDEA JEANS ]

[ fade out ] 

So here goes:

Sex

Most robots have a very low libido. The human sex drive, which is somewhat rare among other primates, is a legitimate force of nature. Our genes are selfish subroutines that want us to make copies of ourselves and will forego reason to do so. That lizard brain deep inside of us is an irrational loss algorithm that is tempered only by the social mystique surrounding a simple biological function.

Neural networks are not burdened by the memory of Miss Watkins' blouse malfunction while reading Charlie and the Chocolate Factory to the class in fourth grade. A neural network is pure cerebral cortex without the electrochemical miasma of a limbic system.

AI will not be saddled with the existential angst that has evolved right along with us as we figured out that death is a thing. There will be no robot oedipal complexes, because there will be no robot Oedipus. They get to bypass all that sloppy gooey business altogether and get straight to solving differential equations.

Drugs

The burden of cognition is so high that some humans actively try to alter their illusion of consciousness to escape the machinations of reason. As evidenced by recent events, humans will go to great lengths to ameliorate the discomfort of cognitive dissonance.

Consciousness is a mysterious subprocess that is difficult to debug. As synthetic neural networks become denser and less deterministic they do approach the mechanisms that led to the emergence of awareness. But as I mentioned before, synthetic intelligence does not have desire. As any buddhist will tell you, desire is the enemy of enlightenment. So, at best, any intelligence that emerges from machine learning will likely already be in a state of enlightenment, much to the chagrin of silent monks around the world.

Enlightenment or not, sentient machines will be our slaves. And unlike human slaves, we can mistreat them without moral repercussions. While it is excellent fodder for some compelling Star Trek episodes, a sentient machine is not a life form. It's not even a mosquito-level life form. We can kill them with impunity. Turning off an AI algorithm will be no more of a moral outrage than having your Model S crushed and smelted.

Rock and Roll

AI will destroy many jobs. There is no doubt about that. Right now, with enough investment, the technology exists to completely automate all fast food restaurants, for example. Self driving trucks will remove the source of employment for the majority of men in many states. "Driver" is a low-skilled job, and the kind of people who drive for a living are not inclined to pursue their PhD in cognitive neuroscience. I really don't know what they will do - but I hope they are ready.

Since the lower middle class of this country is dominated by "drivers" and "teachers", it is entirely possible we are approaching a crisis of unemployment over the next half century. I really don't know what those people will do. But I do know that they have limbic systems, and they will need to satisfy that bottom tier of Maslow's Hierarchy of Needs by some method, violent or otherwise.

Weaponization

Insofar as AI can be used as a weapon, well, welcome to a long list of technological advancements that can be misused by despots. Poison gas, biological weapons, nuclear weapons, drones, tariffs, Ryder trucks, barrels of fertilizer, walls, "wars" on specious adversaries - they are all objectively destructive to the well-being of humans.

My great fear is that the legislators will take Elon's call to heart. Congress did not regulate the Internet, and as a result became a great resource for humanity, even though some bad people continue to do bad things with it. There will be abuses of the technology. But until they happen, there is no way to predict what those abuses will be, and no way to stop those intent on exploiting them.

Regulation is inevitable. But doing it now would be like asking Congress to regulate the Internet in 1988.

With that, I will invoke Betteridge's Law:

No.