Monday, August 13, 2018

Future-Thinking

I often think about the future. "How long is this pizza going to take?!" Just kidding. I do think about the future though and the headline about an MIT computer predicting the end of civilization as we know it in 2040 gives us a good chance to make some predictions. First, let's dismiss the MIT computer.

The MIT computer program was written in the early 1970's and it predicts that based on the out-of-control rise of the earth's population, growing pollution, and the lack of new resources, mankind is doomed. And 2040 is the year of that doom, with our population becoming unsustainable and then crashing, presumably to oblivion.

The problem is, this is all BS. This is as laughable as the movie "Soylent Green" which whines about New York City and the other big cities on the Northeast coast merging into one giant city of 23 million people! It's so crowded that people are constantly touching shoulder to shoulder and everyone is starving! Oh no! But of course, their view of overcrowing is ridiculous. New York has 23.7 million people today and none of this is true. Texas is 171 million acres, so you can literally move everyone in the US onto a half-acre plot each in Texas and leave the rest of the US empty with room to spare in Texas. What's more the world's biggest (and most prosperous) cities are larger than 23 million people today... and nobody's living shoulder to shoulder or starving.

It's the same problem with the MIT computer. According to EPA date, pollution is WAY down from the 1970s. Just to give you a sense, even since 1990:
Carbon Monoxide is down 77%
Lead is down 80%
Nitrogen Dioxide is down 56%
Particulate matter in the air is down 40%
Sulfur Dioxide is down 88%
And they were already way down from the 1970's. The MIT computer seems to think these are going up. Population growth has slowed dramatically and will soon start to decline (I don't know of anywhere that has even a growth-sustaining birthrate). Food, a vital resource, has soared. The World Bank says yearly food production has grown 700% since 1961. Oil production is up 40% since 1980 despite repeated claims of reaching peak oil. Natural gas is up just over 100%. Biofuels didn't even exist in the 1970s. Solar was nothing and is today starting to become significant. We're nowhere near tapped out on Uranium. So basically, every single factor the MIT program relied upon in seeing the end of the world was wrong.

But hey, if you want to panic, feel free. The media loves a good panic.

Anyways, let's talk about the real future. It's going to be interesting... and difficult.

Here's the thing. Believe it or not, there isn't a single job that a machine can't replace. Let's start with the obvious. A vast number of jobs are already gone to things you didn't even notice, like word processors which wiped out the ranks of secretaries and typing pools. Digging and lifting machines wiped out basic labor. Let's call this Phase One of the Robotic Era, shall we? In this age, labor saving devices still operated by humans (a forklift, a word processor, a car) wiped out a vast number of menial jobs.

In Phase Two, those devices become more intelligent. Indeed, as robots get more sophisticated, they are increasingly able to take over all the physical labor jobs that used to require human dexterity and human judgment. Most things can be built by robots now. Kiosks can replace service workers... in fact, the McDonalds ones are snazzy. Soon, you will have self-driving cars wiping out truck drivers, bus drivers and cab drivers. Even surgeons are using surgical machines now. Because we don't trust them yet, most machines still need a human supervisor, but that is what the human is becoming -- a supervisor, not an operator. This is where we are now.

In Phase Three, AI will become capable enough of controlling these machines without a human supervisor. Factories and places with limited human interaction will be automated because there is no need to have a human involved. A smart AI can control the machines better than a person. This means places like banks and fast food and anything that could be run as a kiosk will become automated. Humans will be left just managing operations rather than supervising specific machines.

In Phase Four, more "intelligent" jobs will vanish. A lawyer machine with the proper algorithm and access to a legal database will give better results than all but the most creative humans. Banks, accountants will all be run by AIs, with human control limited to the very top. Think of it this way: a Board of five humans oversees the AI which runs the other AIs which manage the bank's processes and control its branches. There is no need for human involvement. In fact, I think you will find that humans become an inferior choice to using AIs at this phase because AIs have access to greater knowledge and their decisions are less subject to bias, ego or random influence.

This is when things start to go wrong. With no need for humans in most industries, there is no need for employees. In fact, the only ones needed at this point are those with the creativity to give the machines purpose. So what do you do with the rest? With no jobs, do you hand them money? Or does money go away and we enter a sort of machine-based communist era? After all, communism does work when you have enough slaves. I think at this point, you see the end of human society as we know it and the world becomes a guilded cage of welfare. What do the humans do? Well, you say, we do sports, arts, entertainment!

Only, most people can't do that even if they had all the chances in the world to do it. So I suspect most people will simply become little more than what we see in the ranks of billionaire's children today -- worthless hedonists who turn to parties, drugs and other destructive ways to pass time. Sadly, that's what human nature suggests.

Moreover, sports, arts and entertainment won't offer a refuge for long because of Phase Five.

Phase Five is when AI's finally become creative enough that human input is no longer needed to give the machines purpose or let them meet unexpected challenges. This is the end for what's left of human endeavor.

Take sports. I've long wondered why women's sports almost never come near the popularity of men's sports. Is it sexism? I don't think so. I think the answer lies in the pyramid of popularity. If I had to describe the popularity of a sport (let's take soccer), it seems to flow from World Class Event (i.e. Word Cup) to best professional leagues (British, German) to best college leagues to women's leagues to local high schools to kids. What causes this? Talent. Humans seem to be drawn to the groups with the most talent. Sadly, once people start offering robot versions of sports, I think this desire to watch the best will lead to the relegation of humans to second tier and eventual hobbyist only. You could say that humans will always play in college leagues, but do you really need colleges if no one is going to have a job?

In terms of arts and entertainment, the same holds true, and I think CGI has already shown us that "fake" will have an advantage over the much more limited "real." Moreover, as AIs are developed to the point of being able to exercise creativity, they should pretty easily surpass humans in the arts. Indeed, look today and you'll see that films that could have been written by algorithms are dominating the box office and "artistic" films languish. How long this Phase takes to arrive will depend on how long it takes to write creative AIs, but it will arrive. And after that, I see little left for humans to do.
Phase One: Human Operator
Phase Two: Human Supervisor
Phase Three: Human Manager
Phase Four: Only Creative Humans Needed
Phase Five: No Human Needed
Ultimately, I'm not trying to be pessimistic. What this means is that humans will need to find other ways to shape society rather than the "work for a living" model we use today. As policy makers, people need to start considering this too.

Thoughts?

20 comments:

Tennessee Jed said...

Well, it is scary to contemplate, of course. I guess science fiction writers have been imaging this in one form or another as when "terminator" decides no humans are needed and decides to do away with them. While that part is conjecture, I do wonder, how and if, emotions can be built into the alogorythms (sure why not!) your premises seem logical and well thought out. The questions that remain: will the speed of change continue to accelerate? Perhaps more importantly, will some natural disaster (super germs, nuclear warfare, bio-logical terrorism) come along and set technology back far enough we have to start almost over? I am thinking here of the classic survivalists. How many modern humans could survive without today's food and shelter systems? Sometimes, those events are hard to predict or foresee.

tryanmax said...

Man's eyes are constantly on the stars. We will colonize the planets; we've already made our minds up about that. I think terraforming entire planets will extend and expand phases 2 & 3 for a very long time. I don't see how we can hope to colonize the solar system without highly reliable AI-driven robotics to secure oxygen supplies ahead of human arrival.

AndrewPrice said...

Jed, I don't honestly know about emotion. From what I can tell right now, rather than trying to do that, they are simply writing algorithmic approximations of things like emotion. In other words, right now, AI is kind of a myth. It's not independent thought, it's programmed thought that merely looks independent. I suspect that's the best we will ever achieve. BUT that's still enough to bring about what I've written.

Would a natural disaster set us back? Absolutely. It's also likely that roboticization will advance at a vastly different pace throughout the world. And it's also possible people will pass laws putting limits on what robots can do.

That said, I think this is inevitable eventually and we should probably start thinking about how the human race needs to handle the future, lest we become a bunch of mindless billionaire babies living in a robot-run world.

AndrewPrice said...

tryanmax, My daughter jokes about becoming the dictator of Mars. I keep telling her though that her empire will have very few people as it will just make more sense to send armies or robots and only a small number of people until and unless we can truly terraform the place.

And even then, with birthrates as they are, there may not be that many people to spread around the universe in the future. We're becoming more panda-like, which may be how nature puts a "use-by" date on species.

AndrewPrice said...

BTW, This isn't meant to be depressing. This need not turn out bad at all. I just see this as our obvious current path and some of the problems it raises.

tryanmax said...

People also have a propensity for inventing new stuff that they don't need, and in the early phases of development, it's just not cost effective to automate. There's also a market for authenticity. I suspect this market has not reached it's maximum potential.

AndrewPrice said...

tryanmax, I think there will be a market for "human built" for a while, but in my experience, things like that don't appeal to the vast majority of the public.

tryanmax said...

I wondered if I needed to expound on "authenticity." One way to fulfill that market is through human built or made-by-hand things, but it goes much further. Religion sells authenticity, as does science. Talk shows, reality TV, and social media all sell authenticity. Mass produced bottled soft drinks and cheeseburgers sell authenticity. What is authenticity? I'm not entirely sure, but I'm sure there are a lot of counterfeits. How this keeps humans employed, I'm not entirely sure, but it seems to work well and I think it will keep working for a long, long time.

AndrewPrice said...

Let's not jump to conclusions on this bicycle "terrorism" thing in Britain. This could just be normal road rage against bikers. I think it's time we had the courage to ban bicycles. Damn the National Biker Association and their intransigence! It's time for action!

AndrewPrice said...

tryanmax, I think you are talking about two things. The "authenticity" movement seems to be a nostalgia thing, but it usually just aims to whatever the prior generation had. So it might start as a "human made" consumer movement, but it will eventually give way to machine made.

The other thing seems to be some sort of "you are better than the herd" instinct. This is how they can sell mass produced products as "setting you apart" and "making you a leader." That instinct ("I am independent, look at the stuff I buy!") will always remain, but it's easy to manipulate, so I don't see it sustained a human economy for very long.

Art is probably the one thing where a human touch will always be a good selling point, is my guess.

Tennessee Jed said...

I agree on arts .... and maybe it gets to my point in emotion. Let's take Bob Dylan as an example. Dylan's "artistic worth" so to speak, is not through the beauty of his melodies. They are non-complex tunes, but he does tell interesting stories, and though he hardly has what most would consider as a good voice, he sells the story. If you have the time to check it out, there is a song titled "If you see her, say hello" on his album "blood on the Tracks" in which he sings of the bitterness of a break-up but how he has since mellowed and moved on. His voice is cutting. This is where AI might struggle to replicate. It might produce Bach like tunes better than Bach, but cannot help but sound computerized.

ArgentGale said...

Interesting thoughts, Andrew. The later phases remind me a bit of a game I played where there was a society so advanced that work was considered a privilege and everyone else didn't have anything else to do except consume entertainment, primarily through an amusement park or a gigantic virtual reality MMORPG - with said game actually being our universe! Leaving aside the game world as our universe part the concept didn't make sense to me at the time (and still doesn't in some ways now that I actually know what being in the workforce is like) but automation to that extent really would lead to something similar. This got me thinking about how much virtual reality would play in the hedonism you describe. If it gets as good as Star Trek's holodecks or better that's another sizeable can of worms.

That said, I do agree with Jed's point about art. No doubt that those AIs can create things that are technically wonderful but the best stories and songs have heart to them and i don't think that's something that any kind of machine can replicate. All that said, how would you prepare for this, Andrew? It's easy to think of how this situation can go horribly wrong so I'm curious as to how you'd get the good turnout you mentioned.

Unknown said...

It’s all fun and games until someone loses an AI

Critch said...

"Because we don't trust them yet, most machines still need a human supervisor, but that is what the human is becoming -- a supervisor, not an operator. This is where we are now." I love that quote. When I was teaching people how to get a job and keep it I often said the same thing, just not nearly as eloquently.

Doctors used a robot to operate on my oldest son and save his life. Without that machine I don't think he would have survived.

AndrewPrice said...

Critch, It's probably the "da Vinci". It's an amazing machine that is making surgery so much safer -- as is all the automation.

As I said above, none of this is necessarily bad. It all depends on how humans respond to the changes.

AndrewPrice said...

Allena, LOL! Nicely played!

AndrewPrice said...

Jed, What art offers that other fields don't seems to be the ability for the imperfect to become perfect. Hence, some singers with truly awful voices or strange lyrics produce some amazing songs -- things it doesn't seem a machine could create.

But we'll have to see. The "smarter" machines get, the more likely they are to figure this out too. But art does seem to be the one thing that is most likely to remain largely human.

AndrewPrice said...

Hi Daniel,

I'm glad you enjoyed my ponderings. I like to stretch my mind now and then.

I'm not honestly sure how I would respond. I think I would probably pass a law always requiring some level of human supervision to (1) keep the human labor market alive and (2) prevent us from becoming nothing more than chimps sitting in the capsule as the machines do all the work.

I also think I would require a fairly intense level of education, by law, so that people still learn skills and (hopefully) have dreams of progress.

That's probably just a start, but it's all I have at the moment. I'm struggling to think of other ways to keep humans engaged.

Unknown said...

Thanks Andrew,
I too find your ponderings fascinating.
I agree, there would need to be some sort of way to ensure the robots can’t be hacked or EMP’d.
And everyone should know what to do if for some unforseen reason the decepticons go on strike.

AndrewPrice said...

Thanks Allena!

Hacking will be a huge problem, especially when self-driving trucks first become prevalent.

A robot strike would be a disaster! LOL!

Post a Comment