Skip to main content


A Youtuber called axe863 made a video in which he used scientific, mathematical and statistical common-sense to deliver the KO that the Venus Project and Zeitgeist Movement so richly deserved.

If his approach seems weird and unconventional it's because he's not attacking from a tradition neoclassical or Keynesian perspective. Axe863's poison is complexity economics, something a good deal more dangerous to ideas like TVP and TZM. [2]

Now to a couple of comment threads from below the video that I thought could od with being replicated just in case they get deleted at source!


AstralLuminary1 year ago
Why can't we generalize the consumption patterns of middle-income people in the western world, set our constraints equal to the amount of localized resources, and the rate of resource recovery, derive a population growth model that would be sustainable to said consumption patterns, and derive the necessary quantifiable amount of work required to expend a self-sustaining system of sorts

Aistis Raulinaitis1 year ago
Ok, just try the first part. Not even everything else you set forth. Try to generalize consumption patterns.

How could you know a year ago that people would want, for example this thing called the apple watch? How could any program you write possibly ever come up with something as complex and yet desirable as an apple watch? Now lets go through the rest of your claims of trivial calculation, localized resources.

How do you know how much Ullmannite, Bunsenite, and Cooperite you have in the cubic meter of ground underneath you without measuring it? That's a localized resource, it should be trivial to compute, right?

As for the rest of the things you can supposedly "derive", you can see the pattern of difficulty in any one of these calculations. Let alone things like labor and tool inefficiencies, the complexities of cultural differences of quality of labor, quality differences in goods produced leading to unexpected feedback loops in the economy. It's not all just about quantity.

The best you can do is to look at historical data and then try to have some sort of regression analysis. But you might know that good retrodictive capabilities does not imply good predictive capabilities.

Nassim Taleb and many others have spoken about this exact subject in the context of algorithmic trading and how even the HTF guys are still constantly vulnerable to "black swan" like events. Meaning all market actors, even the supposedly evil algotraders [traders using algorithms] are constantly exposed to unpredictable forms of counter-party risk.

Now imagine this situation, the entirety of all human action is orchestrated by the program that you wrote. Think about the kinds of strange interactions we get when we have a large amount of actors exposed to a black swan events in the market in things like flash crashes and even larger events like the housing crisis or the looming sovereign debt crisis.

Even with the resiliency that a market provides, through losses being spread across all the participants, there was still much worry about the stability of our current system. Now think about what would happen if your program encountered one of these events and it was the only actor in which risk could be spread across. It would be disastrous.


I love your thoughts on this. Where is part 2? I am currently in the early stages of engineering an RBE that will hopefully have the following properties:

* Nonoptimality but sufficient use of resources to out compete/subvert threats to the system, and then maintain enough ecological influence to prevent the regrowth of those threats.

* Emerges from the bottom up, starting as a privately owned company.

* Uses the actions and goals of individuals instead of surveys to obtain their preferences on reality.

* Is a synergistic system consisting evolving AI components, a hive mind of those who use the systems, and the partial resources of those who use the system.

* As far as biases present in the individuals using the system, they will be counter acted with real world experiments (see below), and eventually moral, emotional, chemical, cultural and ideological programming of individuals to obtain both a non-homogeneous optimal exploratory pattern of idea space (not every one will think the same) as well as the correction and minimization of biases. The more the system believes something is a bias the more it is minimized via the previously stated programming, however it is never stamped out completely to allow the hive mind to continue exploring idea space in that direction.

* Those who can not afford to pay to use the system may:

1) Carry out tasks in the real world for the system.

2) Carry out experiments in the real world for the system.

3) Verify beliefs of the system.

4) Make positive changes to the system (which go through a series of test beds; rough simulation, peer verification (of those deemed creditable by the system), and larger and larger real world tests before implemented on a macro scale) <-- this is key to the system. There will of course be regional differences in how the system operates, do to cultural and environmental variables that are beyond control of the system, as well as it's constant need for real world experimentation and diversification of risk.

* Because it is a corporation, it will have to provide people with things they want to sustain itself.

* Attempts to become as automated as possible to zeroize monetary costs (free) and minimize labor costs but will start with a cost to get off the ground.

* This system will NOT attempt to destroy or overthrow governments but perhaps eventually mold and absorb them into itself, as will be it's general philosophy with all living agents to maximize ecological harmony, and the well being of all living agents. Anyways, I'm a computer programmer and this is and will be my life's work, but if too many people think of this as a dystopia rather than a great improvement on the current ecology of systems, i will make changes. I have been working on a prototype for some time now and will be ready to release the first product soon. (I'm very curious what you think)

This sounds interesting. At the same time I would refrain from grandiose statements. Hope to see more. There could be an interesting discussion.

I can share a lot information about experience of elite Soviet scietists who worked in the field of resource allocation, some of their papers as well as some more recent papers in the field of optimisation and others related fields, etc. This might lead you in much more realistic and productive directions.

Yes i can get carried away, i'll try to avoid the grandiose statements. What i'm currently building is a human intelligence amplification tool based on game theory, decision theory, and Bayesian statistics.

I will sell this tool to enhance the intelligence and strategic clarity of the individual. Later I plan to incorporate various learning algorithms into this tool such as deep nets and genetic algorithms.

The key to embody all of this is a very fluid and easy to use and understand GUI and interface that prevents the user from being overwhelmed with complexity, confusion, or restriction; but i am not yet ready to publicly share this.

The tool will then be given the ability to share models built with it, such as those models built for fitness function of humans in general or specific types of humans. As the tool gains complexity (hopefully I could afford a small but skilled team at this point), the 'models' being shared will drift towards the flexibility and nature of 'code' or functions that can encompass and use other functions.

In a very smooth transition the system will become a full blown AI, being programmed by the intelligent people of the world, who are feeding resources into it in exchange for its ability to help solve their problems. Things get tricky at this point but it will have to start trading human labor for its solutions, its strategic management of companies, individuals, etc, which most of it at this point would have been programmed by people who were just 'using' the system.

I have a lot of details that still need to be worked out, especially as it nears the later stages but the high level concept is as previously stated but specifically an organic and flexible system that takes off with feedback loops, goes some where at least, and benefits people. I am very interested in your soviet stories and papers. Could you please share? :)

I'm busy so Ill keep this short. I like your name even though gradient descent has a very low rate of convergence. Part 2 will probably never come.

1. How can you even start to quantify the degree of non-optimality without a risk adjusted efficiency measure?

2. "Uses the actions and goals of individuals instead of surveys to obtain their preferences on reality."

Unless otherwise unconstrained individuals internally imposed budget constraints and said budgets constraints were derivable without measurement error, a model would either be biased or have inflated errors depending on whether it is an independent or dependent variable.

3. "Is a synergistic system consisting evolving Ai components, a hive mind of those who use the systems, and the partial resources of those who use the system"

Where does your system even obtain information relevant to sup { Utility of individuals} w.r.t constraints ? How does this synergistic system react to the fact that allocations made by the system through time can change the degree of systemic fragility and agent dynamics?

4. What is the underlying nonlinear stochastic process driving your system? Economists have worked for decades advancing dynamic optimization. Their models fail because they take a non-complex theoretical framework. Any complex adaptive system exhibits a degree of irreducibility and radical unpredictability (non-trivially non-stationary process).

This was well highlighted by Wolfram where he aptly classified evolutionary processes and capitalistic systems as Type IV systems, systems with irregular innovation based structure changing dynamics. Capitalistic systems have demonstrated with sufficiently long sample size sufficient robustness to processes that are orders of magnitude beyond current mathematical representation/statistical estimation.

axe8632 years ago (edited)
Given what you know in computational complexity theory and the foundations of scientific rationalism, what should be the null position concerning the non-empty yet to be quantified high dimensional RBE type optimization problem?

Do you assume, as your null, that it will find a solution in "quick" finite time or that it will not find a solution in "quick" finite time until demonstrated otherwise?

GradientDescent2 years ago

These are good questions. (If you don't have time to read this, basically I thought about it and think an RBE will always be destined for failure) I've decided not to abandon the monetary system.

I skipped question 4 because of my conclusion at the end of this.

RBE Null hypothosies: Actually when i even just start to think about foundational beliefs required for it to work, I can't think of one that's even remotely plausible to be worth testing. RBE may be possible in the distant technological future but even the government welfare/handouts of such an advanced/automated society would be better than an RBE.

Null hypothosies for my system:

1) Less than 1,000 users in the first two years of release, means not enough people are interested in using a system like this, or it provides no sufficient benefit beyond common sense.

2) Less than 20 users who create anything new or beneficial with the system, means those whom the system can afford to incentivize, lack the intelligence to improve the system and those with the intelligence to improve the system, the system cannot afford to incentivize.

Actually, this is almost certain to be a failure point. I'm going to cut the null hypothesis short as it would almost certainly fail. I do very much however want to create a run away adaptive system and i'm going to keep working on my project with that goal in mind, i'm hoping if the system can make me intelligent enough, that will feedback into my ability to enhance it in a loop.

Maybe I could let people have it for free if they allow it to run operations in the background of their computer, so I at least have a supercomputer to experiment with more computationally intensive AI subsystems. I'll leave that a problem to be solved and avoid null hypothesizing it. I still do believe it's possible for every one on earth to have the opportunity to work for food, but it seems like that's more of a governmental problem and as inequality is the engine of progress it may not be best to step so far off the gas pedel... just yet. Hey thanks a bunch for taking the time to post some road signs to get me back to reality. :) I get lost sometimes. lolz. Anyways, going to make a habit of Null Hypothesizes for everything now. Thanks again!

GradientDescent2 years ago (edited)
For posterity's sake, of those who would like us to pull off an RBE, here's 7 RBE pre-requisites off the top of my head. (probably not perfect but it's a starting point to thinking about RBE feasability) Alpha) Voluntary_hours >= Labor_hours_required + education_hours_required

1) Alpha, as required for RBE sustainment

2) Alpha, as required for the creation of automation systems required to lower the right side of the equation, lower than the left. This means, while they are still working for food, unless we make food free in our current system.

3) RBE promises a minimum standard of living for all. So all jobs that currently pay more than affording that minimum standard, will have to sustain a drop in manpower as individuals currently need to be incentivised beyond that basic standard of living to spend a large portion of their lives at work, or in school for those rewards, or will need to be automated.

4) Labor allocation based on the pleasure felt from doing the job (to include a sense of contribution) would have to be sustainable for society.

5) Those who want a better standard of living would have to be discovered, and prevented from restarting a monetary system. If not the monetary system re-evolves, the inequality of the poorer RBE's living a purely RBE lifestyle may push more of them into the labor market until the RBE culture collapses or is marginalized.

6) RBE would have to happen all at once world wide, or whatever country that did it would be taken over by a non-RBE country with a more powerful government, or the RBE country would need protection from the United Nations, with the UN willing to protect that country.

7) For #6 to be possible, the RBE memeplex would have to be capable of infecting the majority of the minds on earth. This might be possible given memes like Christianity, Islam, Agriculture, and Clothing. But would be no easy task.


[1] The Logical Destruction of TVP part 1 by axe863

[2] Complexity economics article on Wikipedia

[3] Desirable Properties of an Ideal Risk Measure in Portfolio Theory by various

[4] Radon-Nikodym Theorem article on Wikipedia

[5] The Radon-Nikodym Theorem by Yongheng Zhang

[6] Pricing in Terms of Market Probabilities: The Radon Nikodym Theorem by Jan Roman

[7] Lecture 10: Change of Measure and the Girsanov Theorem by Steven Lalley


Popular posts from this blog

What Lingos Are Most Similar to English, Though?

Commentaryism - The Death Toll of Capitalism

How many people have died because capitalism exists? How many would still be alive if it had never existed? Let's dig in!

We will take two approaches over the course of this blog post by looking at the the death tolls attributed to the word in its broad popular definition - everything socialists don't like - versus the toll that fits the definition offered previously on this blog.

By the same token I will not lay any outsized figures at any other mode of production's door except where that mode of production demonstrably caused the problem that killed people. It's political ideologies that really matter here, and this is where the first big problem with even trying to lay a specific body count before capitalism runs into problems - there is no political ideology called capitalism.

Now then, Alfonso Gutierrez says in a comment thread that "capitalism and free-markets have murdered billions of people" which is a risky claim at the …

Trickle-down Economics as Economic Theory in Reality

I watched an interview with Deirdre McCloskey on the Youtube channel of the Institute for New Economic Thinking. [1]

After doing so I contributed to a comment thread, recreated in full below, wherein a chappy who claimed to be an economist tried to convince me that trickle-down economics actually is a serious thing after all. This was in response to my posting a link to Thomas Sowell's article The Trickle-Down Lie, and I am so far unconvinced by the tale the economist in question spun for me.

He cited a paper from the 90's as his example, and I entreat you to have a gander at its abstract and compare that to trickle-down as described by David Stockman in his interview with William Greider on supply-side economics. [2][3][4]

Steve Horwitz isn't in love with the phrase, but offers a decent definition;
It’s hard to pin down exactly what that term means, but it seems to be something like the following: “those free market folks believe that if you give tax cuts or subsidies to …