/w=3840,quality=90,fit=scale-down)
Raphael ThysIn my last post, I outlined a realistic but alarming scenario that describes how society could get reorganized due to the impact of AI.
But I also promised to talk about ways to avoid this dark scenario. I am going to talk about that in this post.
- I will first quickly describe the scenario outlined in the previous post again to refresh your memory. (I am going to call it the “Elysium” scenario. The reason for doing so will become clear soon.)
- Then I will give a justification for why the Elysium scenario is not only possible, but even likely. Unless we do something about it, that is.
- Then I will go over why I believe that there is a way to avoid that scenario.
- And finally I will describe some ways to do that.
The “Elysium” Scenario
To refresh your memory, here is the picture from the last post.
Class hierarchy of the future, reorganized due to AI (Image credit: Vin Bhalerao)
This is essentially the class hierarchy pyramid that has always existed in human society, but “stretched” vertically as if it was made of rubber. And I have introduced some additional elements that come into the picture as a result of advances in AI, and automation in general.
This class hierarchy of the future has the 3 well known classes: upper, middle, and lower, or, as described in a somewhat dark but tongue-in-cheek manner in the picture, Sociopaths, Clueless, and Losers.
And I have subdivided the Clueless class further into 10X techies, AI, and 1X techies, with the role of AI expanding and pushing people in the other layers out.
As it turns out, the movie Elysium essentially describes a very similar and quite realistic society of the future.
In it, the “Sociopaths” of the world live in a space city, supported by a huge number of “Losers” on earth, and a few mostly “Clueless” mercenaries in the middle.
Elysium, the space city in the movie, is a perfect embodiment of the idea of the “bubble” I have depicted at the top of the class hierarchy pyramid.
That’s the reason why I call this the “Elysium scenario”.
And since we are in such a Hollywood mood, let me bring in a reference to another great movie.
The Shape of Water
Note that the class hierarchy picture shown above looks almost like the shape that forms at the surface of water when you drop a rock into it. I call it a “liquid pyramid”.
For example:
A “liquid pyramid”. Photo by César Couto on Unsplash
This is not a coincidence. The similarities between this liquid pyramid and the class hierarchy picture go a lot further than just their superficial appearance.
Water forms this peculiar shape because it is liquid, dynamic, flexible. (If we think of molasses instead, we would get a much more regular triangular shaped pyramid.)
The society of the future shares the same characteristics. Due to AI and other technologies, the rate of innovation will be swift, leading to an even more dynamic society than today. Disruptions will occur a lot more frequently.
We are already familiar with the idea that, due to technology, the rate of change or disruption in society has been accelerating. AI will accelerate it even further.
So, if you think of such a dynamic society as if it were a liquid, and every major technological disruption to the economy as a rock that gets thrown into it, then it is easy to see that the result will be a “liquid pyramid” exactly like I have described.
And the similarity doesn’t end there.
Liquid pyramids are short lived
Space city floating above the earth
The first thing you will say when you look at the “liquid pyramid” above is that the droplet at the top is going to fall down, followed by the rest of the pyramid. The whole structure is guaranteed to collapse.
It is my contention that this will hold true for the class hierarchy also. This is because, like I said earlier, the society has many of the characteristics of a liquid.
As an aside, without providing any non-obvious spoilers, let me just tell you that in the Elysium movie, the pyramid doesn’t last even for 2 hours! (Ok, I know, that’s the movie running time, not the actual timeline in the movie, but you get my point. 😊)
Beyond physical similarities, there are logical reasons for this: the Sociopaths live in a bubble where they lose touch with reality since they aren’t experiencing it anymore. They are governed by the laws of the bubble which are made up. We have seen this happen multiple times now with stock market bubbles and crashes.
Plus, the upper classes depend heavily upon the Losers and the Clueless below them, who, over time, become discontent and either protest or cheat or poison or even attack them.
In the old days, things used to move slowly, so this process used to take years or decades or even centuries! But with the acceleration in everything, one can expect this to happen much more quickly in the future.
The high rate of disruptions also imply that the pyramid is constantly threatened. An innovation may disrupt the AI layer itself, or the Clueless people will migrate to a newly forming pyramid caused by a disruption. (We see this phenomenon also occurring already where companies and even countries are constantly trying to woo or retain talented people. This will only accelerate in the future.)
Now, if the pyramid collapses in an uncontrolled manner, it can cause a lot of devastation and chaos. We have seen many examples of this also in history. And it can take a very long time for things to get back to stability and then growth again.
It is in everyone’s interest to figure out how to prevent things reaching this level of extreme so we don’t have such violent collapses.
We already have mechanisms like the Federal Reserve that tries to take the air out of bubbles slowly or take other measures to avoid sudden collapses. Their efforts may not always work perfectly, but we haven’t had a major depression or revolution lately, so I’ll call that progress.
So, what are the ways that we can ensure that we don’t end up with the Elysium scenario?
But before we look at that, you probably have a more basic question: If this accelerating change due to AI has the potential to cause such turmoil, why bother developing AI at all?
Remind me, why are we building AI in the first place?
When human beings tamed fire, they were able to keep themselves warm on cold nights, cook their food, keep predators away, and so on. All of this improved their quality and even span of life drastically.
Unfortunately, the discovery of fire also led to violence, accidents, and later, smokestacks that pollute the atmosphere, even to the point that it now threatens our civilization.
Still, we would never think of going back to a time before we had discovered fire, would we? The benefits far outweigh the costs, provided we manage them well.
The same continues to hold for AI.
The whole reason why we have been building AI (or automation in general) is because it lowers costs, increases our speed and reach, reduces drudgery, improves safety, and generally improves human lives, both at work and at home.
And of course, it also has downsides. The question is whether we can manage them well.
This brings us back to the question of how to prevent the Elysium scenario. What can we do?
This has become a controversial term, but I mean it in a much broader and realistic way than it has been portrayed in the media and I hope you will hear me out.
The concept I am referring to is wealth redistribution.
Wealth redistribution is inevitable, one way or another
Photo by Sharon McCutcheon on Unsplash
It is unfortunate that the word “redistribution” has become politically charged. It need not be. The economy is always redistributing wealth in one way or another. The question is whether it is inevitable and if so, how to do it safely.
And, as I have already described above, the liquid pyramid is inherently unstable, so wealth redistribution is inevitable.
So let me present the various ways it could be done. (Some of these are well known, but others are my own speculations.)
A) Universal Basic Income (UBI)
The most well known way to redistribute wealth is to tax people at the top and redistribute some of the money collected by giving everyone a basic minimum or “living” stipend. This is generally known as Universal Basic Income (UBI).
This could be called the socialistic approach to redistribution.
There is a legitimate concern that this approach may take away the incentive to work or innovate for many people, though many others could actually flourish due to not having the constraints of earning a livelihood and may actually work harder and more creatively.
B) Universal Basic iNvestment (UBN)
The capitalist way would be for the people at the top to realize that investing broadly in the entire society is ultimately in their own best interest (as I have tried to do in my own previous post). I like to call this Universal Basic iNvestment (UBN).
A simple form of this already exists in the form of widely diversified index funds, particularly those that invest in emerging companies or economies.
Another way could be microloans which have proven to be quite promising if used carefully.
Other mechanisms such as crowdfunding or airdrops or other crypto concepts are constantly being invented.
Here there is a legitimate concern are that not all Sociopaths may participate in UBN. Also, the investments may not reach everyone.
C) Universal Basic Payment (UBP)
Another capitalist way would be to pay people for their data, since it was their data that was used to train the AI. The easiest way would be to simply pay everyone some basic amount of money on a regular basis. Maybe we can call it Universal Basic Payment (UBP).
D) Universal Basic Conflict (UBC)
If none of the above are put into practice, then we really end up with the Elysium scenario. In that case, it is possible that people may resort to non-cooperation or sabotage or some other violent and suboptimal methods of making the liquid pyramid collapse.
For the sake of continuing the nomenclature pattern, I am calling this Universal Basic Conflict (UBC).
Historically, this has been the way that hierarchical pyramids that have gone to extremes have crumbled, so this is certainly within the realm of possibility. But I am hoping we have learned our lessons and will hopefully avoid this fate.
We must admit that, at this point, we do not know which of these solutions will work.
My hope is that various parts of the world will experiment with various combinations of these ideas, innovate new ones, and keep adjusting them as needed. We have a lot to learn about how to do this properly.
Still, maybe you have a more basic question: Why would the Sociopaths agree to share their wealth? And what good will come out of it?
I talk about some ideas on that in the next post.