Jump to content
Objectivism Online Forum

Government Run by Computers?

Rate this topic


Andrew Grathwohl

Recommended Posts

If the only legitimate government is one run by the tenants of reason, logic, and truth, then would a computer-run government be more efficient and more free than one run by men and women? (Supposing that computer software were able to calculate - in real time - the extreme stochastic processes involved, and computer hardware were reliable enough). If not, why would Ayn Rand have been against such a government?

Link to comment
Share on other sites

If the only legitimate government is one run by the tenants of reason, logic, and truth, then would a computer-run government be more efficient and more free than one run by men and women? (Supposing that computer software were able to calculate - in real time - the extreme stochastic processes involved, and computer hardware were reliable enough). If not, why would Ayn Rand have been against such a government?

Sure! Sounds perfectly just and fair...so long as I am the one who gets to program it.

Link to comment
Share on other sites

If the only legitimate government is one run by the tenants of reason, logic, and truth, then would a computer-run government be more efficient and more free than one run by men and women?

There's still the matter of programming. There was a popular expression among programmers a few decades back "Garbage in/garbage out."

Link to comment
Share on other sites

More free? You must be referring to the Government’s administrative department, and maybe the court system, only. Otherwise, it’s a very unrealistic scenario, unless you’re talking about T-800 models, at least. Such an advanced technology would make very difficult to finance a small number of police stations, let alone an entire army. This is certainly not what Ayn Rand had in mind when she said the Government should be an impersonal robot.

Link to comment
Share on other sites

Comedic oppurtinities abound....

All power would ultimately rest in the Department of Technical Support (except such opwer as would reside in the Light and Power department).

We'd see news items like this "..of course this war could have been avoided if the president of [country name] had checked his email."

At election time the people could reboot the government.

Government crashes would be a frequent occurence.

There'd be groups unhappy with the government's operating system, whoc'd then set up their own governemnt running their favorite Linux distro (meaning of course their dozens of governments, each running a different distro).

McAffee and Norton would fight pitched battles for the antivirus and firewall contracts.

We'd see obscure conspiracy theories involving a second partition government.

Link to comment
Share on other sites

My question was actually a serious one. A computer can calculate things really well, and operates on a seemingly perfect level of logic, but they cannot possess true moral values. Would Ayn Rand have been comfortable with a government that is derived from moral values (the morals of the people) but whose executives and governors possess no ability to comprehend morals? Would a computer-automated society be of greater use to the theoretical Objectivist government, or could a complex computer system never be able to govern with constant automation due to its inability to be human?

More free? You must be referring to the Government’s administrative department, and maybe the court system, only. Otherwise, it’s a very unrealistic scenario, unless you’re talking about T-800 models, at least. Such an advanced technology would make very difficult to finance a small number of police stations, let alone an entire army. This is certainly not what Ayn Rand had in mind when she said the Government should be an impersonal robot.

But would it be correct to assume that any technology able to accomplish this would be out of everyone's price range (particularly an Objectivist government)? I feel like the free market would be keen in bringing prices down and increasing efficiency, as it always is. Obviously, we would need to factor time into the equation; it would be very unrealistic to expect such technology and intellectual capability to exist at the present time.

Edited by Andrew Grathwohl
Link to comment
Share on other sites

A computer can calculate things really well, and operates on a seemingly perfect level of logic, but they cannot possess true moral values.
Let's check that assumption that a computer operates on a perfect level of logic. A computer is not capable of an emotional response, but it is also not capable of full logic. A computer has no ability to form concepts, which is an important aspect of the faculty of reason. A computer cannot even grasp (even in an operationally-defined sense) ordinary statements in human language. Before you allow a computer to regulate the use of force in society, you need to prove that it does always adhere to objectively-justified moral concepts, and that it is capable of expanding its knowledge to correct errors and address novel contexts.
Link to comment
Share on other sites

Let's check that assumption that a computer operates on a perfect level of logic. A computer is not capable of an emotional response, but it is also not capable of full logic. A computer has no ability to form concepts, which is an important aspect of the faculty of reason. A computer cannot even grasp (even in an operationally-defined sense) ordinary statements in human language. Before you allow a computer to regulate the use of force in society, you need to prove that it does always adhere to objectively-justified moral concepts, and that it is capable of expanding its knowledge to correct errors and address novel contexts.

Perhaps "logic" was the incorrect word. You are correct in saying that logic requires much more than perfect calculation, and right now computers do not have the ability to address moral concepts.

There is no reason why a computer could not form concepts, though. Indeed, it's a level of complexity that nobody has successfully been able to get a computer to calculate, but the potential is still there, isn't it?

Link to comment
Share on other sites

There is no reason why a computer could not form concepts, though. Indeed, it's a level of complexity that nobody has successfully been able to get a computer to calculate, but the potential is still there, isn't it?
I don't see the evidence that a computer can form concepts, nor do I see the evidence that a computer has the potential. Feel free to provide the evidence, but remember that the ability to imagine something does not constitute evidence. Here's a brief sketch of my reason for thinking that computer-run society is impossible. Concept-formation requires free will. Concepts are unique to man, and serve the needs of man's cognitive nature. A computer simply does not have man's cognitive nature. (Economy, for example, is not an issue for a computer). A computer -- an automatic non-volitional error-free calculation device -- is not volitional. Once you're introduced that which is logically a prerequisite to moral judgment, you've introduced that which allows reason to be set aside.
Link to comment
Share on other sites

Let's check that assumption that a computer operates on a perfect level of logic. A computer is not capable of an emotional response, but it is also not capable of full logic. A computer has no ability to form concepts, which is an important aspect of the faculty of reason.

I would go further and say that any consciousness capable of concept formation (and thus logic) must also have emotions, as they are two necessary expressions of the same mental capacity. Before identification, you must have evaluation, and with selectivity comes fallibility.

Artificial consciousness would not make better choices because it is artificial, but it would have the capacity for superior choice because of its capacity for exponential self-improvement.

Link to comment
Share on other sites

How about the technical issue of how principles are refined or expanded? Let's not put ourselves in the problem of a 100% entrenched philosophy that leaves no room whatsoever for improvement. Even setting aside the issue of principles, we'd need a system of laws, as opposed to principles. No computer is going to decide that the appropriate punishment for Crime X is prison/fine/something-else Y. Laws have to come with the consent of the governed, and people are not (and ought not to be) convinced by "the computer said so".

Link to comment
Share on other sites

I don't see the evidence that a computer can form concepts, nor do I see the evidence that a computer has the potential. Feel free to provide the evidence, but remember that the ability to imagine something does not constitute evidence. Here's a brief sketch of my reason for thinking that computer-run society is impossible. Concept-formation requires free will. Concepts are unique to man, and serve the needs of man's cognitive nature. A computer simply does not have man's cognitive nature. (Economy, for example, is not an issue for a computer). A computer -- an automatic non-volitional error-free calculation device -- is not volitional. Once you're introduced that which is logically a prerequisite to moral judgment, you've introduced that which allows reason to be set aside.

A computer has the potential to form concepts for the same reason that a computer can already synthesize audio, receive tracking information from a computer mouse, and detect specific movement within a streaming video. These are functions that could not be performed on a computer from the start. These innovations had to be carried out over time - over the development of the computer itself. Every time a computer innovation comes out, it builds off of something more fundamental and basic. As an example, look at the improvements being made to artificial intelligence, and compare their performance to artificial intelligence from twenty years ago. Computer innovation stems from two continuously changing contexts: 1) human invention; and 2) hardware/software efficiency. Our knowledge of working out problems on computers extends constantly, and because of this, hardware/software capabilities are constantly increased. As long as the code is logical, the code will work - unless the hardware or software cannot handle it. In that case, the inventor may need to also figure out how to design software or hardware to his/her advantage.

That entire process can and likely will provide us with a future filled with computers that operate cognitively. Peikoff wrote an article once about the analytic-synthetic dichtonomy where he stated the two possible answers to any statement about an existant: “X (which means X, the existent, including all its characteristics) is what it is”—or: “X is not what it is.” To me, this is a mirror of the base 2 (binary) numbering system. This same concept is applied to every aspect of a computer, because ultimately, everything a computer does is based off of a single boolean value from which all other functions stem. Ayn Rand narrowed everything down to: Do I want to live, or do I not want to live? As long as there exist only two possible answers to a question, and only one of those two answers is true, there will be the potential for a computer to perform any calculation or task of any level of abstraction - as long as it's logical, where the ultimate and most fundamental boolean value the computer calculates == true.

Edited by Andrew Grathwohl
Link to comment
Share on other sites

A computer has the potential to form concepts for the same reason that a computer can already synthesize audio, receive tracking information from a computer mouse, and detect specific movement within a streaming video.
Your analogy is not apt. If you had shown that a computer can figure out how to synthesize audio or detect specific movement, you might have the start of an argument. (The ability to receive a stream of data is not at all impressive, so I ignore the mouse example).
As long as the code is logical, the code will work - unless the hardware or software cannot handle it.
Which is the point. Why would you even think that the hardware and software of a computer is capable of concept formation. Concept formation is so radically different from what computers do that I can't imagine what fact would suggest that it is a "potential" of a computer.
This same concept is applied to every aspect of a computer, because ultimately, everything a computer does is based off of a single boolean value from which all other functions stem.
There is a difference between saying that you could post hoc model a specific chain of logic involved in concept formation by a man, and saying that you can mechanistically compute concepts that will be formed from observation of reality. Since computers lack free will and free will is essential to concept formation, computers cannot form concepts.

I also want to speak in favor of the relevance of emotions in concept formation, as GC mentioned. Selective evaluation is a sine qua non of concept formation.

Link to comment
Share on other sites

The real question is what are the roles, filled by government, that a computer would need to assume? If the role is keeping a registry of driver's license data - computers are already used. If the role is making and enforcing objective laws - this is a function man is perfectly capable of performing. Government is about power, if a computer did 'derive' rational laws that man is somehow unable to derive himself, then men must choose to enforce those laws. Or, robots could enforce them, but someone would have to initially build them. Humans could even control the robots. The point is that there's really no reason for a computer to be the government.

Unless your concept of government involves compulsion and a centrally planned society - which is obviously immoral. In this case, I can't see how Ayn Rand would support the idea of a society run by computers.

The difference between saying 'society' and 'government' is that society is comprised of individuals who are free to make their own decisions. Government is the institution that can use force to compel. Therefore, it might be rational to use computers with advanced stochastic modeling to make decisions - as individuals, households, firms etc. - but for the government to use such models to regulate and enforce personal and economic decisions is something completely different.

Edited by ZSorenson
Link to comment
Share on other sites

A computer has the potential to form concepts for the same reason that a computer can already synthesize audio, receive tracking information from a computer mouse, and detect specific movement within a streaming video.

This logic does not hold. Just because computers are capable of doing X now does not logically infer that computers might be capable of doing Y in the future.

Edited by RationalBiker
Link to comment
Share on other sites

There really is no argument against computers which are "exactly like humans" performing tasks as well as humans, except the one: they don't exist.

They don't exist yet. I do not wish to argue that they do - I only intend to posit that computers have the potential to operate cognitively.

This logic does not hold. Just because computers are capable of doing X now does not logically infer that computers might be capable of doing Y in the future.

I didn't intend to convey this. My general support for this claim was written beneath that passage, as quoted below:

Ayn Rand narrowed everything down to: Do I want to live, or do I not want to live? As long as there exist only two possible answers to a question, and only one of those two answers is true, there will be the potential for a computer to perform any calculation or task of any level of abstraction - as long as it's logical, where the ultimate and most fundamental boolean value the computer calculates == true.

If you had shown that a computer can figure out how to synthesize audio or detect specific movement, you might have the start of an argument. (The ability to receive a stream of data is not at all impressive, so I ignore the mouse example).

The ability to receive tracking streams was impressive at one point in time, which is why I included it. I'm not old enough to remember the invention of the mouse, but the evidence is there to suggest that the mouse was a pretty inventive and brain-splitting invention for the computer upon its initial release.

Again, I would like to reference my quote above to answer you. I am not a good enough programmer to know how to accomplish such a task as getting a computer to "figure out" things on such a complex level. That would be the work for an incredibly skilled and inventive theoretical computer scientist. What I can say with certainty is that the computer can simulate pretty much any model or form that it its told to unless that "ultimate boolean" == false. Perhaps a model of the human brain could accomplish this task, or maybe there needs to be a massive development of stochastic processes that always return the same results for both humans and computers. I don't know, however I do know that anything that can be expressed mathematically can be abstracted by computer programming.

Why would you even think that the hardware and software of a computer is capable of concept formation. Concept formation is so radically different from what computers do that I can't imagine what fact would suggest that it is a "potential" of a computer.

Well hardware would simply need to be able to calculate the (relatively) extreme amount of processes introduced by such a program. The software would only have to operate to the extent that whatever OS the program is operating on is efficient and versatile enough for such tasks.

Concept formation is not radically different from what computers can do today; concept formation is radically more complex than what computers can do today. If we accept that reality operates ultimately in base two (which we have), then that's all we need as proof that concept formation and cognition are features that a computer has the potential to possess. Had you told a computer programmer in the 1960s that one day computers would render entire feature films using nothing more than complex vector and raster synthesis, and animate the whole thing using automation, he/she probably would have believed you, but would not have had the slightest clue as to how that would work. My point is that extremely complex developments happen over time, and we cannot be expected to be able to conceptualize the process for accomplishing these tasks - only the groundwork from which these tasks will be developed. If we understand the operational basics of a computer, then we can understand its limitations.

There is a difference between saying that you could post hoc model a specific chain of logic involved in concept formation by a man, and saying that you can mechanistically compute concepts that will be formed from observation of reality. Since computers lack free will and free will is essential to concept formation, computers cannot form concepts.

This is true - as I've said already, I am making a distinction between what I can do on a computer presently and what somebody else can potentially do on a computer at some point in time.

Free will, as it is defined with regard to humans, is something that can never happen on a computer because it simply executes commands. I'm not sure how that is related to computers engaging in concept formation, though. Computers already engage in concept formation on simplistic levels - that's the basis of machine learning. I would imagine that this technique will only improve in quality and begin to mirror human learning in the years to come, due to the massive usefulness of such an ability.

I also want to speak in favor of the relevance of emotions in concept formation, as GC mentioned. Selective evaluation is a sine qua non of concept formation.

This is a legitimate point. Should a capitalist governing body operate on emotion?

Edited by Andrew Grathwohl
Link to comment
Share on other sites

They tried this in Terminator and it got out of hand. However, particular functions of the military and policing are becoming more and more automatized--particularly surveillance and security related issues. If you asked me whether I'm in favor of computers calculating the trajectory of missiles rather than keeping a handful of mathematicians awake and ready for war all night, my chips would land on the side of the computers.

Link to comment
Share on other sites

  • 9 months later...

Have any of you heard of Project Cybersyn? It would be interesting to get your thoughts on this in preparation for an essay: http://en.wikipedia.org/wiki/Project_Cybersyn

You should read on the fate of the Allende government. His central planning destroyed the economy so rapidly that he was overthrown in a coup three years later.

Link to comment
Share on other sites

  • 11 months later...

There really is no argument against computers which are "exactly like humans" performing tasks as well as humans, except the one: they don't exist.

It appears we may be on our way, though!

http://www.computing.co.uk/ctg/news/2102735/ibm-unveils-chips-mimic-human-brain

Link to comment
Share on other sites

  • 2 years later...

I think it's a fascinating idea, with serious potential, provided we acknowledge one fundamental thing.

This computer must, in some way, be completely dependent on human beings (unless it's self aware, in which case it would be functionally identical to conventional governments anyway).

So no matter how advanced this nonconscious computer is, someone will have to build it, someone will have to program it, someone will have to repair it and give it regular maitenance; et cetera.

So as far as that goes, you can't really have a totally automated government; somewhere in the process actual human beings will have to intervene.

Which means that we would then have to figure out what the best computer-assisted governmental structure would be, and that's what I find so fascinating.

Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

Loading...
  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...