I seem to recall that some economists believe that central planning of an economy is an NP-hard or even hypercomputational problem, meaning that rational economic planning may not be practical even with AIs. From the abstract of the linked paper:

Quote

The standard view of the socialist calculation debate is that Mises and Hayek at best demonstrated the  practical impossibility of socialist economy, but that the mathematical solution of economists such as Dickinson showed that “in principle” planners could achieve a rational use of resources without private ownership of the means of production. In the present paper I hoped to show that this view is incorrect, because(if seriously implemented) a socialist planning board would need to publish a list containing an uncountably infinite number of prices.As Cantor’s diagonal argument from set theory shows, it is demonstrably impossible to construct such a list.Therefore, socialist economy is truly impossible, in every sense of the word.

This other paper would seem to suggest otherwise, presenting counter-arguments. Anyone with more knowledge of complexity theory than I want to weigh on in this?

If a centrally-planned economy is computationally infeasible, how does that the change the ideas presented this thread for dealing with "AI-conomics?"

12 hours ago, deltaKshatriya said:

I don't believe banning automation is really a solution either, since you can't stop the entire world from not using it, and, ultimately, is just a means to kill progress.

1) Is progress always the highest goal to strive for? Is progress always a net positive for humanity as a whole? Or has the blind faith in the good of science and progress become a substitute for religion, where people cling to the believe that progress can cure all ills given time, and good things await beyond the "event horizon" (whith death being replaced by the far future here).

One reason I never called myself an atheist is because I cannot really get into this blind faith in science.

2) You can't stop the whole world. True. Just as much as you cannot make the whole world adopt your algorithm driven system. In all probability every part of the world will react differently to automation... leading to a new competition of systems and, maybe, cold war in the process.

Some parts of the world will profit a lot, others will loose in the process. Leading to more inequality over the whole globe, no less.

So if we want to get back to reality, no "solution" will work everywhere, and no "solution" will be generally adopted, and thus every utopia you think of right now will help a region become more equal and prosper, at most, not the whole world.

12 hours ago, deltaKshatriya said:

I'm more curious about this. How would we offset the job loss from automation in this system, however?

I think the most realistic base idea has already been put forward, the basic income. It was even voted on here in switzerland, and while it had predictably little chance of being accepted, it achieved double digit yes votes... I think something in the 30 percentage region. Many young people, and many in my age group where open to it (I voted yes btw)....

And the most amazing thing was that even a lot of professors of economy, and even some right lobbyist actually had a positive stance to it, calling it one of the few leftwing ideas in the last few years that were actually good ideas for the economy.

Thus I think a general basic income has a good chance to be accepted by all parts of society, given enough time. There are good selling points for everyone, being able to sustain the poor with a monetary source of income that come without the stigma of other social wellfare programs because everyone gets it, and being a leftwing idea going in the direction of socialism will satisfy the left, and it having the ability to soften the blow of automation and all the efficiency and money saving programs of the private sector in the last few years, AND reduce the cost of the wellfare program considerably makes it attractive for the right.

Now, will that alone be able to carry a society that has a very low employment rate? At that point, it comes down to the productivity of the economy and high taxes for private corporations. Will companys accept these far higher taxes? That is a good question. 

Sure, socialism and communism, where companys are owned by the state might make this question redundant. But I would say that given the taxes are not punishing in the sense that companys do not see any profit in fully automating their business instead of just paying less efficient human workforce and foregoe the taxes (because IF taxes are too high, someone somewhere on the planet will employ low cost human labour and just produce the things the oldfashioned way.... or produce it the automated way WITHOUT paying any kind of taxes), probably companies will accept the higher taxes given these taxes guarantee a stable system and social peace.

After all, what we have seen in the last few decades is that stability is just as, if not more important than efficiency or low cost. A lot of companys moving their labour to low cost countries have made a net loss because those countries lacked a stable system... or because they lacked regulation for a stable work environment.

I guess give it 10 more years with the current cluster**** with globalization and outsourcing, and some companies might be happy to pay more for a "quality work environment" in a stable country.

Again, note that this is not a complete deviation from the more extreme system you are proposing. All I am saying is it can work WITHOUT moving to a fully state-owned economy.

Maybe a question you should think about is this:

- How would your system interact with other systems implemented in other parts of the world?

It probably would be naive to simply assume everyone would find the same solution for the same problem at the same time. It would probably also naive to assume this algorithm driven socialistic/communist system would be the winner of this new "race of systems". So you could face the situation were you have another competition of systems, like during the cold war.

How would your system react to that? As far as I can tell, communist systems always had problems with global trade, as their system was built upon a closed ecosystem, thus outside interference would probably mess up the plan, and thus the system.

Now, you can say that throwing AI at the problem might make a "five year plan" unnecessary, the plan could evolve from second to second, thus making the system more robust against outside interference, thus making trade with other countries/systems/entities not an automatic threat to the system.

But lets consider this: if, for some reason, a competing system (ultra-capitalist, or another communist system, doesn't matter) can produce what your system is struggling with cheaply and offers you a good deal. Should your system now take advantage of that to fill a shortterm need? Even if that means longterm that you are loosing value, as you are now paying someone outside of your system for a product, thus you have a net loss, while your system is based on a closed system, as far as I can tell? Or do you close your system to outside trade, in a way like China did for a long time, and still does in some way, or what Trump wants to do in the US?

Can your system survive without protectionism?

13 hours ago, deltaKshatriya said:

The only major thing you cannot own is factories/means of production.

Seeing how this is actually the biggest "power" in any system... how to mitigate the risk of abuse of centralizing so much power? An incredibly big government like China has (where the senat has thousand of members)? A more complex government, were there is no central authority, but instead give the separate entities "ruling" over parts of the system (the heads of state owned factories, the different departments) more authority?

14 hours ago, deltaKshatriya said:

1): We don't need to have necessarily a central AI. This can be a conglomeration of AIs talking to one another, where each AI represents some sector of the market. It depends more on how the system is designed. 

2): Capitalism does have some key dependencies as it is. The Federal Reserve Bank has to function, for example. Banking in general has to function. If banking fails, capitalism fails right with it. Determined actors can easily take this down. Imagine someone who infiltrates the Fed, for example.

3): Capitalism is abused quite a bit as it is. As @Luckless pointed out. Capitalism isn't very robust to abuse either. We had the Great Depression, after which we realized that more fail safes are needed. Is it tough to imagine people building a similar system failsafes for our proposed system here?

1) Good. Then we also need to make sure these AIs cannot easely be cracked with the same key... thus they are working from different location, have radically different code, different keys, and so on.

I do think such a system probably could be pretty robust, if its designed for security and being hard to abuse... but its going to be a delicate balance between a lot of different interests

2) IDK if capitalism really IS that dependend on banking. You could happily replace banking with an unregulated system like the cryptocurrencies and capitalism could still thrive, probably. Not that I think deregulation is going in the right direction, but after all banks are just a middle man. The federal reserve is important as long as we want real life value being tied to our currency, and we trust states more than private entities... something the cryptocurrency craze has proven is not a given anymore, altough I personally take the conservative stance when it comes to cryptocurrencies to "wait and see"... most of it at the moment is hype, and when the dust settles, some few people will have made a fortune, most will have lost a lot of money, and if cryptocurrencies really will live on hinges probably on wheter a currency comes along that actually offers something existing solutions cannot provide, without coming with a ton of downsides on its own.

3) Capitalism IS abuse. But it has proven a very stable system, having survived quite a lot of crises... Mostly thanks to most states still having some "socialist" components to them, and not fully privatizing the system. And the fail safes you mention.

Capitalism taken to its extreme would probably lead to anarchy and fail quickly... whereas communism taken to its extreme leads to stagnation and will fail just as quickly.

So yeah, I am sure we have a quite wide spectrum to move on where we could develop a stable system. Its not on the extreme ends of the spectrum, but certainly a little bit more socialist ideas couldn't do much harm to our current capitalistic systems.

14 hours ago, deltaKshatriya said:

IMHO, space colonization is at least a couple centuries away. That and I'm still not very sure if people will WANT to colonize space, given how little value we might get from it, at least, for now.

Well, the question is do people want equity, really? Devils advocate here, can't resist

The system you propose here hinges on enough people actually WANTING to change the status quo, which also isn't a given. And how people will react when sh*t hits the fan, and the system has to be changed because of a detoriating standard of living, is still up for debate. So if we debate a communist system that probably will not have a chance in real life because of a multitude of factors outside of how well designed the system is and how well it would work (could be the best system in the world, and then people vote for Trump, or Erdogan)... I am sure we can debate a space exploration program that is unlikely to kick into high gear in the next 100 years... both for the small chance that thing move quicker than we anticipate, and for the sake of internet discussions

As to the value of what could be out there... we simply don't know for sure. The POSSIBILITY for value is endless, though. Helium3 is just the beginning... how many rare resources restrict new technologies because we simply cannot not source enough of it here on earth? What other stuff could be out there we don't even know about yet?

The cost is astronomical (pun intended ), sure.... the benefits are not clear, sure... but as long as we stay a capitalist system, its not a question IF it will be done... just how much cost will have to come down before a company is willing to take the risk.

Then there is human curiosity. If the last few decades have shown us one thing, its that the more people become atheists (or simply non-religious), the more science becomes a substitute for religions. Thus ANYTHING that can give humanity meaning will be undertaken, including the search for knowledge. Space is kind of the final frontier here.

Again, its not a question IF, but simply by WHEN it becomes cheap enough...

Its a stretch to see a big increase in efforts here in the next 50 years, but at some point humanity will ramp up their efforts to explore space big time. Maybe human workforce is already obsolete by then, and will not even be considered... but human workers might actually have one big advantage over most machines when it comes to space exploration... being "jack of all trades, master of none", while most machines will always be purpose built for a special task. Its one of the reason WHY machines are so much more efficient at that one task than a human worker.

When you send an exploration force into space, you have plan for the unexpected. You probably will not be able to build a machine that can handle the unexpected as well as a well trained human astronaut for quite a long time into the future. Thus this MIGHT be a new venue for increasing demand for human labour.

7 hours ago, mikeman said:

I can't speak about mathematics, just a simple observation/question : In capitalism, the free market is supposed to act as a giant "diffused" calculator that assigns prices to each item, correct? Obviously this "calculator"(composed of ~6 billion human brains atm) has some finite information and computational power. How could it be *impossible* for a central computation to "replicate" this? I was the one that mentioned Mises in the first place, but it always seemed to me the most crucial problem is how to get true(or as true as possible) information about the wants and needs of the population into that central planning body.

If perfect rational planning really is computationally intractable, then it seems obvious to me that the free market wouldn't be able to handle it, either. Relatedly, I would point out that there's a reason we have "market disruption" as a term. No doubt some would argue that the free market not being able to handle it is the entire point of moving to socialism.

Or maybe the fact that the free market is inherently decentralized and parallel is the reason it works better - perhaps rational planning can be done in polynomial time, but that polynomial has a huge exponent on it that makes the problem only tractable when computed on millions or billions of processors in parallel.

Or some combination of both; neither a decentralized free market nor centralized rational planning can perfectly plan an economy, but each attain a different local maximum of efficiency. A free market can have some of its "computations" of what people want and need made closer to the point where its source data is taken, meaning that the data is more accurate at the point of computation, resulting in a local maximum closer to the actual optimum point.

6 hours ago, Gian-Reto said:

One reason I never called myself an atheist is because I cannot really get into this blind faith in science.

Slightly offtopic, but I would point out that atheism and "having a blind faith in science" have nothing to do with one another. Even if science was about faith (which it isn't), there are atheists who reject science. Atheist only means "does not believe in any gods." There are atheists who believe in crystal healing, psychic phenomena, and homeopathy. I would bet that there's at least one atheist flat-earther, too.  Besides obvious existence proofs, if you're having blind faith in the things science tells you, then you're not thinking scientifically at all.

Both "atheism = religion" and "science = religion" are memes with no basis in fact.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

1): Yes, progress is a goal to strive for, since we are talking about making life better and understanding the world better.

But is it more important than, say, the wellbeing of people in the here and now? Given progress has to be bought with people having to suffer (being robbed of your job, for example) to enable a future that MIGHT be better than the present, is it the right choice to force people to endure shortterm suffering for a distant longterm gain?

a) making life better for whom? Again, devils advocate here, but you should probably think who will really profit from it, and who will not. And how to make sure MOST people get something out of it, and how to convince them that its in their best interest.

b) Its a goal to strive for... but so is preservation of what you already have.... its the duality of progressives vs. conservatives. Mark that I didn't pose the question "Is progress a goal to strive for?"... instead I was asking "Is it always the highest goal to strive for?".

I guess your answer to that will come down to your personality type and expierience, which can be compressed into the neat little box of "political orientation" -> A conservative person might actually agree with you that progress is a goal to strive for... but that the cost in the here and now is simply too high.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

being scientific is not blind faith at all.

Not what I said. I said that I have problem with the blind faith in science some people, especially in the atheist community, have shown in the past, and present.

Trying to prove god doesn't exist for example... its about as productive as trying to touch your nose with your tongue. Its probably impossible to do, and even if you finally achieve the needed dexterity, its utterly pointless.

I am all for the scientific approach... just be aware that

a) Science is usually just working with a model of the real world, thus prone to errors due to the "resolution" of the model used

b) Science is very limited to the limits of our knowledge. It will be still be limited in 100, 1000, or a million years, if humanity makes it that far

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

2): I'm simply arguing that banning something is about as useless as it gets. Obviously the world will not agree and not all adopt my proposal. I'm not expecting the world to do so. But banning new technology simply because it'll put people out of work en masse is a lot like blowing off your foot with a shotgun because your foot hurt. Really what needs to be done is to understand how automation should be used in our society and how our society needs to change to accommodate the new tech. We didn't ban factories and whatnot during the industrial revolution: we simply changed our societies. People tried to ban factories etc, but that was not the solution then, it isn't now.

I might agree... but the question in the end is, is humanity really ready to live with the consequences of new technology? Why are we now trying to ban nuclear energy in many parts of the world?

Again, I am pretty sure nuclear energy will live on outside of some western countries that can afford to be picky where their energy comes from. Still, you see... it actually took 2 big reactor accidents, and some serious concerns about the ability of the companies running some of the oldest plants in the world to keep these plants save, to molbilize a good portion of the population of the western world to demand a "ban" (if you can call it that, given it takes decades to switch off a nuclear power plant) of nuclear energy... and it looks like at least here in western europe, the "ban" is actually a done deal.

Now that is not much that has gone wrong. While the two incidents have MASSIVE longterm effects on the regions affected, both incidents can clearly be attributed to a failure of the company involved... one was a very stupid architecture involving combustible cooling elements, one was a company simply not prepared to spend a little bit more for the worst imaginable catastrophe to happen in the region because "that will never happen".

All it took for popular opinion to shift was two incidents that simply showed that nuclear energy couldn't be entrusted to a capitalist company that would always try to cheapen out on security. But instead of accepting that nuclear energy wasn't as cheap as it was made out to be unless another catastrophe should keep happening every 20 or so years, and asking for stricter legislation while accepting a slightly higher price, the reaction was "ban it nao!"

So while I don't think a ban is the right way to go forward... its one of the more realistic scenarios should enough incidents with algorithms and AI happen when this technology ramps up. And they will happen, we know it. And the conservatives will be all over it, unless they profit from the technology directly themselves.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

I think UBI is a good first step for the time being. But as automation and technology get better, job loss will only get larger. Beyond a point, UBI won't be enough really...

Though an argument can be made that the UBI would ultimately morph into what I'm proposing anyways.

That needs to be seen, really.

In the end, its simply a way to make sure unemployed people get the minimum they need in a world were employment is no longer guaranteed. I would expect the state to make adjustment when more and more people are affected, as not doing so would pose a problem for the economy due to shrinking consumation of goods.

True on the last point. Maybe society is moving in this direction anyway... and if its an organic, longterm change, it'll probably happen without too much political noise. It will probably also look much different and probably not be nearly as utopist as we might hope it to be, given its probably shaped by the people who are in power today, and are profiting from the current system.... they will make damn sure they are profiting from the new one.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

What you're saying is that they'll want something of value in return. That's not necessarily tough to do either. It's a question of trade, after all.

Right... the question is how the system can cope with trade. As said, what you have outlayed till now is a closed system. Everything is highly controlled. You probably also want to highly control the contact with other systems -> thus isolationist tendencies, limited trade with other systems.

It depends on how the system is designed, really... but expierience tells me somehow that a small dose of capitalism at that point would probably be in order, looking at how communist countries that started to trade with the outside world successfully often started to adopt partial capitalist systems. China comes to mind, and so does North Korea which seems to have created a special zone with partially capitalist rules just for the limited trade it had with South Korea, while that lasted.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

System design is a big aspect of this. I believe a system can be designed that's robust enough. We already have pretty robust systems IMHO that are almost as mission critical. I'm still not seeing how this is as big a concern as it is honestly. Most of these concerns already exist in some shape or form.

Sure, as long as you are willing to compromise, I don't think its an issue. Most of the woes of real life communsit experiments came from dogmatic following the pure theory (or MAYBE martially misinterpreting what MArx wrote, taking some of his words to literally )... with a little bit of compromise, you probably would have gotten a way more robust system at the expense of some centralized power (if Mao wasn't able to dictate agricultural rules, or the central government wasn't expected to make plans for 5 full years!).

So the BIG question for your communist thought experiment probably is this: will the system be designed by reasonable people of diverse outlook, including rather conservative ones, which are ready to admit their ideas don't work out in real life before shtf? Or will it be ideologes that are more interested in seeing their vision realized than actually listening to people and adjusting their vision if needed?

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

Our current system is extremely dependent on banking. The Federal Reserve is a huge part of the banking system. If the Fed goes absolutely batshit tomorrow for some reason, our system will absolutely collapse overnight. This is a bit of a fallacy to assume that our system is not vulnerable. It absolutely is vulnerable.

Sure, because most of the money no longer is backed by gold. Because our financial system has gone crazy quite a while ago.

Bitcoins, while being coined as the cure, are just another crazy invention that go in the direction of these "virtual money" trends, not backed by any real value, and volantile as ****.

The Federal reserve is that important because of that, AND because of global trade. When the state can no longer control their own currency because some speculative bastards at the other end of the world can bet against their currency, you kinda need a defense mechanism.

In this sense, you are right that the federal reserve, at least, is very important.

The banks on the other hand.... well....

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

I mean yea, it has survived, but mainly because we figured out how to best run it over the years. It very nearly did collapse in the 1930s, when people figured out what more was needed. We almost had some major problems in 2008. We still have major problems with exploitation.

While communist systems went stale quickly, and died (or almost died) a slow death because the rigid system would no longer promote the right people to the right job... or only when the stars aligned (somebody had the right family/friends, was doing the right thing, AND was actually competent for the job).

Lets call it a draw, and rather think about how this affects our robot communism. Given you no longer need the right guy at the right place because we now have expert programs for that, the biggest question remains what if somebody has the "key to the kingdom" and can mess with the algorithm? Or are we now pretending that there isn't an elite sitting at the top of this new communist pyramid, not matter how flat it hopefully is, trying to manipulate everyone, and this case the algorithms, to do their bidding? What are the chances no such elite emerges in ANY system, really?

If we don't pretend they don't exist, what happens when Neo-Mao dislikes that the algorithm for agriculture is perpetuating capitalist stereotypes and reprograms it to only plant... I don't know... non edible plants? Kill all sparrows because he hates them? Inadvertetly kill all bees? How do you safeguard your system from tampering by the elite, while still giving enough control to humans to actually be able to trust in the system?

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

also, as a side not, it's funny to think that people becoming lazy is a problem but progress might not be the best thing...  

Well... I don't care about lazy people, as long as I don't have to pay for them... or the state pays them, and pays me also so I can be lazy too. I will use my lazy time to improve my game dev skills while they can lay at the beach, for all I care

And I certainly think progress is good...

But society atm sees it otherwise.... I am merely playing devils advocate here. Because the best intended plan for a new economy will fail if its not robust enough to withstand being picked apart by the conservatives, and society as a whole.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

I can't visualize every aspect of a system and how it'll offset some particular problem or edge case. Obviously I haven't designed a full proposal nor do I plan to propose it tomorrow. 

I'm not saying there won't be problems. There'll be problems, but no worse than the problems we already deal with. Just different ones.

Fair enough. I'll try to contain my urges to pick apart the system a little bit more from now on. Because that probably wasn't the intent of this Thread

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

Moreover, I'm not proposing that this mass automation is going to happen now, or in the next 5-10 years. It's going to be probably 30-40 years when we start to see the first of the REAL major effects. The decades after will probably get into the bigger effects. It's a problem not too far off. 

Okay, that sounds reasonable... I would go up to 50 years, but really, that is nitpicking.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

There's a difference between exploring space and asking for a better standard of living.

Well... see, again, not my opinion really, but WHO is asking for better standards of living?

Certainly not the people actually in power. And the people that could actually USE a better living standard here in the west seem to be barking up a different tree. The mere mention of communism would probably get a lot of lower class guys into a hissy fit.

The third world countries? Sure. Question is, what power do they have? There has been some progress made here thanks to quite left and liberal governments in the west but... will this trend go on?

So while I might agree that there is a moral difference between the two... there is also a difference in money-making capability. One is a huge layout with a potential for an even bigger payout... the other is still a huge cost, with only potentially better social stability as a payout.

When trying to see the world from the eyes of people in power, better not put morality first. People do not get promoted to these positions because the put their morals first, not even on the left.

On 31.1.2018 at 12:18 AM, deltaKshatriya said:

Right now, most people don't really care honestly.

SpaceX and similar companies wasting millions on trying to make space flight cheaper might disagree. Just because the tech hype of the day has moved doesn't mean that there is still quite some interest in space flight.

I hear the chinese really are pumping quite some money into that.