It’s hard to go a few days without seeing yet another article about the impending doom of artificial intelligence. This doom is related to two arenas-
- Skynet
- Taking jobs
Most AI researchers don’t take 1. seriously. Pedro Domingos, in The Master Algorithm, says
“The chances that an AI equipped with the Master Algorithm will take over the world are zero.”
“People worry that computers will get too smart and take over the world, but the real problem is that they’re too stupid and they’ve already taken over the world.”
Moving past the contradiction: that computers -and subsequently AI- have already “taken over the world” and are already causing issues, yet, as they take over the world more, we shouldn’t concern ourselves with them causing more issues? Just ignore the infamous flash crash, how Google started mistakenly identifying gorillas as black people, or how Microsoft created a twitter account which did this:
Nor do you need to take over the world to cause harm.
What AI researchers do concern themselves with though is the jobs issue. Sort of. Here is Domingos again,
“In the meantime, as the boundary between automatable and non-automatable jobs advances across the economic landscape, what we’ll likely see is unemployment creeping up, downward pressure on the wages of more and more professions, and increasing rewards for the fewer and fewer that can’t yet be automated…The transition will be tumultuous, but thanks to democracy it will have a happy ending…When the unemployment rate rises above 50 percent, or even before, attitudes about redestribution will radically change.”
I recommend The Master Algorithm as a way to have a pulse on this stuff, even if this statement sounds insane. “Life will be rough until a majority is in the rough, who will then vote themselves out of the rough.”
- Tumultuous doesn’t get the picture across in this scenario, considering what happened during the Great Depression when unemployment was only 25%. Endless riots, a civil war / revolution / Bane from Batman would break out if something like this occurred. Not exactly a reassuring perspective.
- The 1%, hell the .01%, dominate wealth in America. Yet the NINETY NINE percent still can’t get money from them. In the modern United States having money is more valuable than having a vote. (This has only gotten worse the last few years.)
- Having the majority might not matter-
- Even if you have the votes, you can lose! See: 2016 presidential election.
- South Africa is like 90% black, yet that Apartheid thing happened.
“Eventually, we’ll start talking about the employment rate instead of the unemployment one and reducing it will be seen as a sign of progress. (‘The U.S. is falling behind. Our employment rate is still 23 percent.’)”
It makes sense to sell AI this way. If your job depends on it, then sell it as a coming utopia. That hopefully quells fears and gives you more grant money. It’s like when Silicon Valley sells us the idea of them changing the world for the better, yet in their own backyard San Francisco has 6,700 people living on the street (which has grown of late), with the second most amount of homeless people per square mile in the United States. Guess they aren’t all on Facebook yet…
Because shouldn’t this utopia have happened already?
Why, once farming was largely automated, didn’t we do this already? Why did we invent new jobs? Why did we invent modern work? Why didn’t everybody just say “Look, we have enough food and water, we have shelter, let’s just hang out all day and sleep, eat, poop and bang?” (WHY ARE THERE STILL MILLIONS OF FARMERS???)
Because we humans, like all animals, have to have a means of ranking one another. So we invent ways of achieving status.
–
There is a documentary called “All Work, All Play.” It’s about the rise of esports, which is professional video gaming. To be clear, this is people making enough money from playing games because other people will watch them play.
There is a point in the film where someone goes, paraphrasingly,
“It sounds crazy, but is it any crazier than watching someone try to hit a ball into a few inch wide hole, from a few hundred yards away?”
No, it’s not. Nor is it crazy that while AI can already play video games, that’s not something people are filling arenas to see. People want to see people.
Golf, riding a bike, basketball, football, hockey, baseball, CURLING, the fact these are “jobs” would sound ludicrous to those farmers in the 1790s.
“Uh, um, how, WHAT THE FUCK ARE YOU DOING? WE GAVE YOU ALL THIS LEISURE TIME AND YOU’RE DOING WHAT TO YOURSELVES? WHY ARE YOU SO DEPRESSED?”
(United States, number one baby!)
There is another documentary called “Valley Uprising.” It’s about people who climb rocks. People who eventually started making money from doing so. For. Climbing. Rocks.
I once went on a zip line tour in Vancouver, Canada. At the top of the mountain the sarcastic tour guide goes “Ah, yes, and if you look to your left you’ll see a bunch of water, then there are some leaves, and to our right we have rocks. Breathtaking!”
CrossFit has a documentary and a yearly “Games.” Kind of like their Super Bowl. We pay to watch people lift weights now. We’ve been paying to watch people flex their muscles on a stage for a while.
Again, imagine this in 1790.
“You pay money to watch people unnecessarily expend calories?!?”
Travis Pastrana did this for the hell of it,
And nobody wants to see a machine do that. The possibility of failure -human error- is a big part of what makes these things entertaining. Nobody wants to watch a basketball game where every shot goes in. AI can drive a car around a loop perfectly, but perfect is boring in NASCAR, and explosions are more dramatic if a person is in the car.
Sports -entertainment- has endless permutations. We’re not going to give them up because of AI. If anything, we’re going to push them further. Hence, the bagillion amount of TV shows right now. Or, how in that “Valley Uprising” documentary you see each generation pushes more than the previous, and make more money than the previous. “Oh, you climbed with a partner? I’m going solo.” “Oh, you climbed with a rope attachment? I’m climbing with nothing but chalk.”
Virtually anything can be turned into a sport. Like throwing ping pong balls into red solo cups. Yes, there is a world series of beer pong, just like there is a world series of poker.
People make millions of dollars to talk about throwing a ball on television / radio / podcasts. They don’t even play the game. They talk about people who play the game.
Board games will become:
Artificial intelligence isn’t going to extinguish this. The whole point of these activities is to assess other humans ( / be part of a community). It’s not to help us survive in a “I’m going to starve” / “Being the best farmer means you produce the most food means we have the best chance of surviving” kind of way. If that were our goal, why wouldn’t people stop working once they e.g. have all the money they could ever need? Like A list celebrities? Because that’s not the only reason they’re doing the activity.
(If you’re really interested in this, check out some books about evolutionary psychology. Highly recommend The Mating Mind.)
AI researchers should already know this. We made a computer better at chess twenty years ago. Yet people are still playing chess. Professionally.
We’re not all going to be sitting around singing kumbaya in the near future because of AI, just like we didn’t start doing that once tractors came around. We will continue to find ways, no matter how arbitrary or random, to separate ourselves from one another. Survival of the fittest is not only a food, shelter, water deal. It’s a sexual selection process as well.
–
Let’s close with some non-sport examples- Once recording devices came out, why would anybody ever see live music again? Because we’re human!
What’s a show we’ve heard endlessly about the last year? Hamilton. Broadway is about as primitive as entertainment can be these days.
Nordstrom sold out of their rock in leather product this year. Feel free to look it up, but your imagination already has the correct image. It is a rock. In leather. For $85.
We’re animals. We do weird shit. Macaques steal one another’s babies for social rank. We throw ping pong balls randomly, garnering a hundred million views:
People think of these worlds. People create jobs. Not machines. Machines may create opportunity for new jobs, but people imagine them. The rate at which this happens is the real area for concern, but longterm, AI will neither take every job nor give us endless leisure time.
Meanwhile, Adobe creating a program where it can doctor someone’s voice into saying anything? AI researchers working on transforming a video to be something besides what it is?
Yeah, fuck off, that’s something to worry about.
–
Posted on July 24, 2017