Would You Be Comfortable Being Ruled By An AI Government?

in #technology2 years ago


I have often talked about the development of AI and the rapid exponential progress that we are witnessing and the impact that might have on human life in the future. Nobody knows for sure how this will all play out.

Predictions range from highly pessimistic to highly optimistic and I think that like with most other things, the reality will fall somewhere in the middle. Meaning, there will be plenty of benefits we'll enjoy but there will also be risks that would haunt us and it will ultimately depend on the decisions we take from here on.

The reason why this topic is so significant is because, unlike our other inventions, this particular one has the potential to replace its very inventors. We are literally creating an intelligence that has the potential to become far more intelligent and superior to us and so diminish our role in the way the world works.

The most prominent and far reaching example would be that of governments being replaced by AI. This idea is not a pessimistic one where AI forcefully takes over control, but one where we humans decide for ourselves that AI would be far better at governing than people.

The Artificially Intelligent Government


You might have read about this idea in some sci-fi novels. The basic idea is that some day, AI gets so intelligent and excellent at decision making, that we humans replace the governments of the world with AI and it starts doing everything that a government does, but far more efficiently, at a fraction of the cost and without corruption.

Fundamentally, that would be a good thing in my opinion. We all know the insane number of things that is wrong with governments of the world today and to have an AI whose only objective would be the proper governing and progress of the people, would be such a welcome change.

But on the flip side, we would in essence be under its control. Yes, it would be kind of voluntary (for the greater good) but it would be control nonetheless. In that case, I wonder how many people would be comfortable and wouldn't give two thoughts about it and how many would constantly feel fearful or agitated by a fact that they were ruled by a software.

I also wonder what such a country would be called. Will it make the concept of democracy obsolete as there would be no leaders to choose? Also, there would most likely be a team of humans to constantly monitor and maintain the functionality of the AI system. Does it make them the ones in actual control?

And then there is always the risk of someone corrupting the AI or it realising itself, that we humans are the enemy and need to be wiped off. Being a government would make its task a lot easier in that case! Anyways, what do you feel about the whole concept?


Not really. As AI is ultimately created by PEOPLE who already impose their own ideologies upon it the moment they start programming in the initial rules. Now, if we can assume those people are 100% benevolent and wouldn't have typical human failings that might not be a bad thing. Then again if we could find people like that to design government we wouldn't really need AI. The only benefit of AI is that in theory it would remain true to the initial "constitution".

People often forget that when people make excuses for algorithms, or AI they tend to not pay attention to the fact it was people who created those algorithms, and AI. It was people that tweaked those algorithms and AI.

Any AI would likely have a backdoor to permit tweaking to keep it from running amok. That would be the stated reason. That backdoor could also be used to corrupt it.

The only benefit of AI in this case is it's speed of processing. As tools I think it is inevitable AI will increasingly become a tool used by governments and many other things. Yet I do not see it replacing humans anytime soon.

It is great at certain things. It is atrocious at others, mainly because we don't understand all things in the mind well enough to replicate them in computers yet. We can make simulations that APPEAR to be those things, but they end up typically not even being close to the real thing. Just like a maniquin they give us something to focus on that our brain recognizes and our imagination can fill in the gaps without actually being equivalent to that which they are supposed to represent.

"... AI will increasingly become a tool used by governments and many other things."

Most of those other things will be individuals, IMHO. Government and the most powerful corporations in the world cannot keep us from just sharing silly songs. They sure can't keep us from doing the same thing with AI - which is just another kind of data, after all. AI may be the best example of how technology increases the power of individuals relative to institutions. The power of an institution is that many people contribute some of their power to one endeavor. Many AIs cannot be wielded by an institution or it will lose it's institutional advantage completely over individuals, each of which can employ an AI, or some several, just like an institution.

For this reason, I expect AI to contribute far more to human freedom than the technical tyranny fearmongers prattle about.

Most of what people call AI today, I consider to actually just be what are known as Expert Systems. I've been a student of AI since the mid-1980s. I've watched a lot change. (especially the meaning of words)

What I'd call true AI doesn't exist yet.

We have some amazing Expert Systems though. Systems designed to handle specific tasks, or solve very specific problems. They are amazing at it. As such they are useful tools. Though I don't consider that intelligence.

We have simulations that mimic tasks that we do with our intelligence. Yet they lack the awareness to go outside of the task, and in reality are quite limited.

Though that doesn't mean they are without danger. They can potentially be even more dangerous if one considers they are not able to see without the constraints of the task they have been assigned to.

These are all reasons I expect AI to be most useful to people individually, rather than suddenly able to control us due to it's superior abilities. Expert systems are really useful for specific tasks, like managing logistics, feeding schedules for agriculture, or any other dronelike task, but have practically zero relevance to intuitively managing concatenations of systems in diverse cultural milieus.

Institutions fund the development of technologies, and as technology advances, the specific advantage of the given technology is of reduced effect in extent and duration, and also disperses throughout the population of individuals at ever increasing rates, where the increase in power the new technology avails increases the power of individuals relative to institutions. The more tech advances, the more this power differential decreases, and the faster it does so. AI is liable to be extraordinarily potent, and since it is just data, to disperse faster than physical gadgets.

The benefits to individuals AI brings is enormous compared to the degree of benefits it brings institutions. Institutions have traditionally been able to concatenate the efforts of multiple individuals, and this has delivered to them power that individuals could not match. Since AI enables automation to multiply the efforts of individuals, it is particularly empowering of individuals while relatively far less so to institutions.

You make some great points and I have aluded to the possibility of humans behind the AI being the real people in control and that is surely a scary thought. As you said, AI works based on the initial rules which could be changed by the people who designed it or are monitoring it.

Rather than some corrupt humans?

Humans are unfit to rule. We have animal drives and they dictate our behavior. When we form bureaucratic systems, it is only a matter of time before they become rotten from corruption. It is fairly common to get drunk from power. It's addictive. Humans are much more dangerous to other humans than some AI that lacks any human drives.

AI will be increasingly used in management because of explosively growing data that is impossible to make use of manually. The term AI is pure glorification at this point. We're really talking about machine learning applied to optimization problems.

I'm thoroughly disgusted at corrupt humans abusing power. I would like nothing more than to see every lazy, stupid, corrupt and self-important manager or leader get sacked. Algorithms do not have any these qualities by default. For example, hiring decisions in the government should be delegated to AI. All good old boy's networks doling out their own cushy government positions should be totally annihilated!

Nearly every country has such networks of parasites draining its resources. Every single one belonging to those networks should be fired immediately.

That kind of disgust towards human corruption is exactly why I wrote this article and why a part of me wants that we should one day get a true AI that could replace governments for good.

An alternative to that is government by people but through something like Tau chain facilitating consensus building by scaling up discussions to include millions of people. Such a system should include some very important safeguards built in against oppression of minorities by majorities.

I would be very interested in blockchain paying a role in governance in the future. Great thought!

Blockchain makes cooking the books much more difficult than using a some sort of private centralized ledger even if it gets regularly audited. A blockchain is immutable and should be made public if it is used to track government money.

Exactly. That's why I hope it is implemented in a major way soon.

We don't have true AI. We have expert systems. Algorithms designed to be very good at very specific tasks. Because these tasks can often be things that typically those we consider very intelligent in society typically do (math, calculations, statistics, etc) we consider the machine intelligent.

Though it isn't. It is constrained to the limits of the algorithm.

The important thing to remember is that it is not truly intelligent. It works within the confines of the algorithms as programmed by humans. The CORRUPT humans you are referring to impose their concepts upon the recipe that ends up being the algorithm.

Though unlike a human, an algorithm does not change its mind. So if there is something to exploit or something that ends up being a very bad idea, the algorithm now acting as the government will not be able to see it. Who changes it?

We don't have anything remotely close to true AI right now. We have a lot of expert systems. They are good at specific tasks, but they themselves are not good at things outside of that task.

We don't have true AI. We have expert systems.

I think that's what I meant with:

AI will be increasingly used in management because of explosively growing data that is impossible to make use of manually. The term AI is pure glorification at this point. We're really talking about machine learning applied to optimization problems.

You wrote:

The important thing to remember is that it is not truly intelligent. It works within the confines of the algorithms as programmed by humans. The CORRUPT humans you are referring to impose their concepts upon the recipe that ends up being the algorithm.

Data-driven decision making can make backroom deals somewhat more difficult than total freedom. We will have increasingly complicated decisions made by AI in an independent fashion.

I agree with your disgust at corruption. I note that we enable such corruption by tolerating government. I would also point out that there are many, many more networks of parasites than there are countries, more than there are governments at any level, more than there are corporations and transnational institutions in the world.

We are being provided technical means of producing our own goods and services today at an ever increasing rate, across every field of human endeavor, and it is the parasitic drain of our wealth by industry that centralizes power and fuels corruption.

Join the decentralization revolution, and keep your money. That's all it takes to fire the corrupt bastards profiting off your work - just keep it to yourself, and stop giving it to them.

To listen to the audio version of this article click on the play image.

Brought to you by @tts. If you find it useful please consider upvoting this reply.

That is scary asf! I fear it's gonna reach a point where robots might just take over!How can I reach you on discord? :)

Yeah, but wouldn't it be scary if we voluntary allow AI to control us

Nothing Surprises me anymore when it comes to humanity... That is very likely to happen!

I appreciate that folks are considering this issue. At least we have a considerable amount of time to give it thought before it might eventuate. While further technical development is necessary before AI would be capable of imposing edicts on people, it is not that which will take the most time. It is the extant powers that be. If they are able to prevent handing off their power, they will. Power provides them benefits they will not relinquish voluntarily IMHO.

I am fairly certain that long before AI assumes dominance in human affairs government will be obsolete anyway. Power is gained by money today. There are folks that will claim 'democracy!', but they are wrong. Stanford proved pretty convincingly that voters have not a whit of say in how power of government is wielded just a few years ago, and despite the strong evidence of that research and much else we can observe about the world, ideologues will not abandon their claims just because of facts and reason. So, I consider that power is gained by money today, and that such accumulations of money will become useless not long from today.

Across history, and every industry, we see that technology has always provided individuals greater power relative to institutions. Basically institutions are able to gain power from multiple individuals contributing, but those individuals do not contribute all their power to the institution, as they are loathe to simply become chattel. So, every increase in our ability delivered by technological advance decreases the power of institutions relative to individuals.

Of late centralization waxed ever more powerful, as institutions grew and parasitism on the commerce conducted by individuals delivered great concentrations of money, and thus power. However, we have today passed peak centralization, and those concentrations of wealth are on the wane. Folks will point out that the digits representing money possessed by institutions and their masters are actually growing, but money alone is not wealth. Money is a veil behind which actual wealth is concealed.

Actual wealth is some things money can buy (which is why having money increases your wealth), and many it cannot. A breathable atmosphere, enough food of good quality, family and friends, etc. These things are not increasing for the moneyed rich. Indeed, their quality of life is decreasing along with ours in many of these metrics. While they may grow the digits of their accounts, money itself is becoming obsolete, because of technological advance across all industries.

What we observe is that decentralization is the cutting edge of every field of research, from agriculture to astrophysics. Decentralization is actual freedom, and centralization is actual tyranny, and as decentralization continues to increase, it will deliver more means of production to individuals (which is how it reduces their subjection to tyranny). As individuals profit more by manufacturing their own goods and services, many via AI, the opportunities to parasitize their commerce are fewer, and concentrations of money receive less rent.

Absent preventing technology from advancing - which is done today, to the degree it is possible and profitable to those whose power is at risk - this will proceed to it's logical end, sooner or later. Our extinction is the only thing that can prevent technological advance completely, and thus does not matter to human rulers. They seek to delay possession of important means of production from spreading faster than they can gain control of people's minds, and thereby to maintain their power.

That will fail, IMHO, because of AI. One of the easiest kinds of technology to spread is digital communications, or information technology, and AI is just that. One of the least mature technologies is psychology, and the poor grasp of consciousness, and neurology itself, makes it very unlikely that such a high degree of mental control of free people will mature to the point we can be prevented from acting at our sole option prior to our possession of AI that will enable us to counter it.

I expect my grandchildren will be truly free, because they will make themselves those products that today are slave collars binding us in servitude to would be overlords, if not my children, or even myself. I am old, however, and ain't counting on much for me.

The last observation on this matter I will make is that I am sovereign, and cannot be made to not be (save for some psychological control technology, which is very far off). No one can cause me to lift my finger against my will. They can convince me to do so, but that is not the same thing. Only I can cause it to lift itself. Only I can choose to effect my will, and this is not severable from me. I barely tolerate human tyrants. I'm not going to tolerate toaster tyrants.


You got a 61.24% upvote from @ocdb courtesy of @sauravrungta! :)

@ocdb is a non-profit bidbot for whitelisted Steemians, current max bid is 35 SBD and the equivalent amount in STEEM.
Check our website https://thegoodwhales.io/ for the whitelist, queue and delegation info. Join our Discord channel for more information.

If you like what @ocd does, consider voting for ocd-witness through SteemConnect or on the Steemit Witnesses page. :)

Hi, could you support post on upcoming event, Ai conf - we will be a first time speakers with Ai and blockchain topic, 27 apr:

Transparent and safe artificial intelligence https://steemit.com/blockchain/@gromozeka/graphgraila..
We also will cover federated learning, decentralized blockchain marketplaces, differential privacy, direct data governance.