Elon Musk’s AI predictions should terrify us

How will AI destroy humanity? Will it simply go house to house in robot form, slaughtering us where it finds us? Or will it instead discover that a certain property of our livers or spleens is the most cost-effective form of lubrication for one of its less important robotic joints, and harvest us for that property, as we now harvest chickens, in battery farms? It’s a fun thought experiment, no?

Perhaps it will find us entertaining, in the same way we find the base animals with whom we share the planet entertaining – my children’s hamster, for example, which for their enjoyment in the evenings I allow to gambol in our sitting room inside a little plastic ball. On that basis, it may let us live out our days in relative peace.

If the robots don’t mercilessly kill us, what will we do all day?

Equally possible, I suppose, is that it will quite reasonably deduce humans can be made to work incredibly hard for nothing more than the avoidance of terrific pain – in a way that machines simply won’t be made to work. As such, it could consign us to labor camps where we’ll toil until we drop. At this stage, all options must be considered to be on the table.

By the sound of it, we’ll find out soon enough. Elon Musk – our species’ leading AI pioneer and very possibly its cleverest individual — has lately taken to using high-profile speaking opportunities to make absolutely terrifying prognostications about humanity’s increasingly short-term prospects.

Asked this week in a televised interview with Senator Ted Cruz how likely is the prospect of humanity being annihilated by killer robots, Musk replied: “20 percent likely, maybe 10 percent… in five to ten years.”

Was he joking? Sadly not. Last October, Musk told the assembled corporate overlords at the Future Investment Initiative (FII) conference in Riyadh – or Davos in the Desert, as it has come to be known – that there is a good chance humanity is right now conspiring in the creation of an AI-defined future that will be truly dystopian. “There’s some chance, which could be 10 to 20 percent, that it goes bad. The chance is not zero,” he said.

More recently, he told podcast king Joe Rogan:

I always thought AI was going to be way smarter than humans and an existential risk, and that’s turning out to be true… I think we are trending toward having something that is smarter than any human – smarter than the smartest human – by next year, or a couple of years. There is a level beyond that, which is smarter than all humans combined, which is frankly around 2029 or 2030.

What a staggering statement: the unleashing in just the next five years of an intelligence so seismic – deepening at an exponentially rapid rate – as to be beyond human comprehension. “It’s getting ten times better per year,” Musk said dispassionately at the FII conference. “Four years from now, that would mean 10,000 times better. Maybe 100,000.” Oh, goody. “I think by 2040, probably there are more humanoid robots than there are people… there’ll be at least ten billion,” he added.

Of course, it’s not all doom and gloom – or at least that’s not how Musk wants us to think of what’s coming rapidly down the track. “You could look at it like the glass is 80 to 90 percent full. Meaning 80 percent likely we will have extreme prosperity for all,” he told Cruz.

Yes, an enormous number of people will very soon be made redundant, on the basis AI will do their jobs much faster and better, but according to Musk:

Goods and services will become close to free, so it is not as though people will be wanting in terms of goods and services… you will have tens of billions of robots. They will make you anything, or provide you any service you want, for next to nothing. It is not that people will have a lower standard of living, they will have a much higher standard of living.

Hmm. How exactly has that worked out for the indigenous populations of, say, Australia or the United States, displaced as they were rapidly by Whitey with his far superior education and technological capability? Not brilliantly, you’d have to say, on balance.

If the robots don’t mercilessly kill us, what will we do all day when they have replaced us at our workstations? Will we collectively – all nine billion of us – take up diverting hobbies, for example, baking and crafting, or will we en masse hit the bottle? “The challenge will be fulfillment. How do you derive fulfillment in life?” Musk mused to Cruz. It’s a good question.

A better – and more obvious – one would have been to ask Musk how he can possibly apportion levels of probability to how the AI revolution plays out for mankind. If one AI – by the end of the decade! – will be more intelligent than all humanity combined, then obviously none of us, not even Musk, will be able to second guess it, now or ever. Very evidently, people will be as influential as apes are currently, or woodlice, to the direction of the future. Our species, and all of our collective achievements, will count for close to zero.

But we had a good run. Humans have been around for about 300,000 years, and we’ve been planetary top dog, in terms of brains, for at least 50,000 of them. Now, either a one-in-five chance we are very soon doomed – likely in ways we can’t currently imagine – or else shortly none of us will ever again have to lift a finger thanks to armies of robots that are both cleverer than us by factors impossible for our relatively tiny minds to comprehend, and happy on our behalf to do all the heavy lifting. What could possibly go wrong?

Personally – and not that it will matter when very soon the robot comes knocking – I’m with Tucker Carlson, who last summer said:

I don’t know why we aren’t having any of these conversations…. If we agree that the outcome is bad, specifically that it is bad for people… then we should strangle AI in its crib right now and blow up the data centers…. If it’s going to become a threat to people and humanity and life, then we have a moral obligation to murder it immediately.

Quite.