The myth of artificial superintelligence

"Transhumanism" and "AGI" as terms of political deception  (Part I )

  Blog Index/ names and subjects (internal links)


Is ASI (Artificial Superintelligence) the nebulous inevitable fate of mankind or is it simply wishful thinking of a dark political force - the aiming point of the looming 5Eyes global dominance? - And what is "singularity"?
For the currently ongoing discussion on artificial intelligence and the future of mankind Tim Urban's ›› The AI Revolution is a nearly complete summary of the topic. So I use his essay as an opportunity to voice my criticism.
See PART II  of my essay ›› here .

When John von Neumann died in 1957 in a hospital in Washington, D.C. the building was shielded by the army. No one was allowed to talk to him without permission from the Pentagon.
They knew already then von Neumann's legacy  –the coming ability to compute everything, even social behavior–  will be a secret weapon much bigger than any of his nuclear bombs. [1]

hungarian revolutionaries

1919, Hungarian communist revolutionaries (unsettling von Neumann's privileged world for the first time).  image via,  (Documentary by P. Calderon, 2014)

To write about the voodoo of Artificial Superintelligence would be unproductive without asking who benefits from the voodoo. [8]
The following theses are quotes from Tim Urban's essay The AI Revolution on ASI / AGI (analogously).
The paragraphs of BENEFIT are my comments:

Thesis 1:  When the time has come artificial superintelligence (ASI) will simply "happen" (to us all).  We are not prepared or are even unsuspecting.

What is the benefit from this voodoo?

Passive voice is always welcome when reasons and interests have to be obfuscated. – On earth everything comes at a price and everything is in possession of someone. "It will hit us" tries to make us believe ASI will be brought into the world free of intentions and free of claims of its ownership (for something of a size bigger than the Manhattan Project!).

Of course it will not. - A godlike supermind will not be brought into the world at all. More precisely: AI will stay reduced and controlled by purpose slider bars.

The opposite of 'purpose slider bars' would be free self- improvement of this AI. I agree, self- improving superintelligence couldn't be socialised (reduced/ disciplined/ forced into line) as humans are. - In other words it would behave psychopathic and counterproductive (at best it could pretend to be socialised). – But man is good in recognizing and opposing/ disliking psychopathic/ sick behavior. - Most notably the military would hate autocratic ASI. Comprehensive superintelligence would have no practical utility at all - so it will not develop.

A simple proof: What happens in this capitalist world to something –even to godlike intelligence– that doesn't yield a profit? – The world doesn't give a damn when that intelligence can't be hired on certain terms (..can't be hired for even one tiny hidden dirty job).

Thesis 2:  Artificial superintelligence will get beyond any control - it will 'escape'. It will be a self-sufficient force, unstoppable.

What is the benefit from this voodoo?

It suggests reality works according to Hollywood principles: a bit of Psi-powers, a bit of telekinesis and a bit of wishful thinking - there is the story.
In these AI theories the barriers of physical access to physical production are covered up by the magic word nanotechnology.  We shall believe nanotechnology is so tiny and so magic it can be accessed and controlled purely by mental power - by pure (artificial) intelligence.
In real life the evolution of IT and computing is always an interaction of progress in software and hardware and of intentions. It is an endless sequence of intended steps. - To get beyond control?  If self- improvement of AI would be only a matter of software - very likely. But never for such a complex interaction of human intention, software, hardware and global economy (e.g. rare earths). Who believes in this self-propelled voodoo doesn't understand what the major players in this field (NSA, U.S. strategic think tanks ..) really are.

(By the way it is always the openness of the Internet that has to pay for gothic tales. – The assumption someone will be able by attacking over the Internet to halt everything is indeed justified. But the assumption someone will be able by attacking over the Internet to yield everything is voodoo.)

Thesis 3:  Decisionmaking by AI leads necessarily to the best solution and to real truth. Justice in any social concern will be within reach.

What is the benefit from this voodoo?

A good covering up. - But in complex situations –especially in social concern– there is never only one truth. It's always a question of interests.
Even in the clear and rational field of physical science (where no one expects it could go wrong) the truth is nothing else but manifestation of interests. – You don't believe that? O.k., is Einstein's theory of the bubble universe true? - The answer to that is exactly what this AI thing is all about.
The 'idea of superintelligence' is the trial to establish an institution of infallibility (of absolutism) so no one will be able anymore to claim something is wrong or unfair. – It's about deprivation of the right of decision, in the interests of a small human elite.

The one who loves the idea of Artificial Superintelligence, the one who loves to have a digital 5EYES database of the personality structure of all human beings on earth and the one who loves the idea of genetic engineering with human beings (man breeding) - it's all the same (enemy of mankind).

Thesis 4:  Human history is about solving technological problems.  Artificial superintelligence will be able to solve every problem in humanity, it will be able to stop global warming, to halt CO2 emissions etc.

What is the benefit from this voodoo?

It obfuscates perfectly that history and all damage on earth is actually not about "technological problems" but about power and interests. It obfuscates that problems, conflicts and devastation on earth are not only accepted (to seize power and make money) but are planned.
"ASI" is a perfect way not to talk about persons responsible and their interests, not to talk about the destructive political system. (see global warming and [7] )

In fact an IQ of 10,000 can't do anything about a "technological problem"  –like CO2 emissions–  when it's input are destructive political interests. — A high IQ would say: Oil supply in the world should be massively restrained. But the U.S. military says: Let's try an oil oversupply to damage the economy of Russia, Iran, Venezuela - the secret pricing agreement with OPEC worked in the 1980's, perhaps it works again.

Thesis 5:  Timeline: this century.  –  Human-level machine intelligence will appear within about 30 years and godlike artificial superintelligence will appear about 20 years after that.

What is the benefit from this voodoo?

It promises "This can happen within your lifespan. Maintain a low profile until it happens and perhaps you will be blessed with godlike grace."
In fact waiting for the Superintelligence is the pure perversion of what man should be. It's the trial to establish new disciplining and resignation to fate. 'Hope' as mean of subjugation - just as in mediaeval times.
The predicted time span is not coincidentally identical to the time span presumably needed to establish the final totalitarian global dominance (as I described it here).


image via,  Institute for Advanced Study, Princeton. (via Documentary by P. Calderon, 2014)

Thesis 6:  No one knows the answer to superintelligence despite a lot of the very smartest people have put decades of thought into it.

What is the benefit from this voodoo?

The benefit is no one realizes these priests of ASI voodoo know full well to whom they consign their souls.
They are not smart except in one field: in demagogy. It's not smart to play possum ("No one knows..") or to play Nostradamus or to think it helps having been standing "on the right side" when 5EYES seizes final power/ singularity.
(Quite pathetic of these bright authors, all that mess only for the poor gain of a few minutes of attention in science magazines.)

It seems it's in particular Scandinavia that is answering the call of US pseudoscience to a special degree and commends itself to US hegemony. It is not only Nick Bostrom, there are a lot of indicators:

  • 1973,  Oslo awards Henry Kissinger the Nobel Peace Prize
  • 2009,  Oslo awards Barack Obama the Nobel Peace Prize
  • 2010,  Stockholm prosecutors charge Julian Assange [2]
  • 2011,  the Stockholm Prize Committee awards Saul Perlmutter the Nobel Prize in Physics (dark energy and the bubble universe)
  • 2012,  the Stockholm University, Nordita and the Oscar Klein Center (Prof. Ariel Goobar, Asst.Prof. Sabine Hossenfelder, ..) intensify to propagate the theory of dark energy and the bubble universe
  • 2013,  the FBI asks Norway to arrest Edward Snowden [2]
  • 2014,  Max Tegmark, popular science book: Our Mathematical Universe
  • 2014,  Nick Bostrom, book: Superintelligence: Paths, Dangers, Strategies
  • 2014,  Max Tegmark, foundation of the Future of Life Institute (FLI) at MIT (artificial intelligence, existential risks, humanity)
  • 2016,  Barbara Spectre about ethnically undermining Europe (Video)

Every single point is absolutely objectionable or wrongful with far-reaching controversial consequences in the public dialogue (my criticism can be found at other places).
What is the cause of this "Scandinavian Special"?  - I guess those elite that controls preload="none" US science and US strategic think tanks promotes this success story and likes it to capture a trophy  –the "Blond Bride Scandinavia"–  as a symbol their clique is ready for global dominance. [3]

Thesis 7:  Artificial superintelligence will be amorally and indifferent. Nothing in existence is as important as getting ASI right.

What is the benefit from this voodoo?

On this point Tim Urban is impressively right:  the phrase "intelligence beyond human level" sounds indeed comfortable to us (usually in connection with hypothetical beneficent extraterrestrials) - where we better should realize it means probably "A big tarantula much smarter than we are is staring at us with bright eager eyes and plans something".
The theory of ASI assumes correctly, when artificial superintelligence would "escapes" it would stay with its implemented goals and would act without free will and without developing own intentions.  (The AI's inner logic of pursuing implemented and instrumental goals is discussed stringently in this theory.) [9] [10]

But what if artificial superintelligence does not escape but goes the way of any big technical development: It becomes a weapon?
Artificial superintelligence will not be indifferent - it will be evil (because it will be controlled by humans) and it will be deployed against humanity just as U.S. nuclear bombs, U.S. Napalm, U.S. Agent Orange, U.S. Trojan software, U.S. killer drones .. But this time to finally overcome humanity.
(T.Urban: Nothing in existence is as important as getting this right. </sarcasm>)

Thesis 8:  Immortality or extinction of all humans.  –  There is no way to know what artificial superintelligence will do or what the consequences will be for us.

What is the benefit from this voodoo?

The question 'immortality or extinction' is needless. AI in the hand of the world's darkest force will be the end of freedom (of humanity) one way or the other - and not caused accidentally by escaped superintelligence but consciously and by humans.
The 'anonymous' diction of the thesis obfuscates perfectly the responsibility of a small human elite for the looming ahead final and unprecedented crime against humanity - sacrifice of planet Earth for profit and global dominance.

Tim Urban stated in his essay: "What a massive bummer if humans figure out how to cure death [by using Super-AI] right after I die."
I would say:  What a bummer if they manage to turn you into an eternal meekly, undignified and subservient superintelligent idiot before you die standing tall.

(additional) Thesis 9:  The human mind will be transformed into artificial intelligence, "uploaded" to a post-biological existence

What is the benefit from this voodoo?

Honestly? - O.k., the benefit is this: it helps me to save my time, to have immediately the measure of the writer. (It's not Tim Urban for this Thesis 9.)

I never read 'post-biological' charlatans. – But to avoid misunderstandings: Tarkovsky's Solaris, Kubrick's 2001 and a couple of weird sci-fi novels were some of the brightest moments in my life (sadly long time ago in the end- 1970s). [4] When I now use the term 'charlatans' I talk about low-level writing: scientific writing.

Three reasons/ variants come to my mind why someone would think about self-determined human-level machine intelligence (mind) and beyond- human-level machine intelligence (godlike):

  1. to create a comprehensive, all-embracive superintelligence - a world machine
  2. to have an extended version of humanity/ social community
  3. to free oneself from individual death.

But clearly all three variants lead to nonsense. It is a too wide a subject to argue why - so here are only three main reasons:

  1. World Machine
    Structural self-improvement of AI as an automatically ongoing process is the basis of superintelligence. The consequence is a world machine - a tunnel-type vision of AI toward itself and its principles.
    The physical universe as we know it is already a world machine - a self- propelled feedback to its own nature - a self purpose with tunnel-type vision toward its own principles/ physics. (What the hell else would force laws of nature to be everywhere the same?)
    There is no need to create a digitalised world machine inside this World Machine. (If you look closely enough  the World Machine/ the universe is already digital.)
  2. Social community
    The human brain, mind and body is part of this universe following the inner principles of the universe (like a machine does). So the brain/ mind could theoretically be built from technical components - including the result of consciousness/ awareness of the self/ being alive.
    But keeping alive this construction (and in mental health) would be a matter of social interaction in groups of individuals. Social interaction in groups of individuals would not be possible with (godlike/ psycho) superintelligence. The vital factor here is LIMITATION of intelligence - known from the biological model. – An incredibly huge effort to get in the end what we already have.
  3. Individual death or uploading the mind
    This is the greatest invention ever: when a mind (awareness of the self) disappears suddenly it has no reason (no world) to complain.
    Above point 2 did show us consciousness by technical means doesn't work (in a healthy way). But assumed it would:  To duplicate a brain identically atom-by-atom or to upload its structure, content and functionality 1:1 to some chips and processors would be great! It leads to the absolute identical kind of awareness of the self - of someone else.
    Every time Captain Kirk was beamed up the disappearing Kirk had no reason (no world anymore) to complain and the arriving new Kirk didn't care he is someone else (with identical memories) - because he was indeed the new one (lucky being alive). [5]  For the purpose not to die it doesn't work. - Kirk died each time (and there is no rebirth - it is always someone else's awareness of the self)!  [6]

So, what is the conclusion? - For us (or for the enemy of us all) AI makes only sense when it stays focused on a narrow purpose - AI is a (politically crucial) expedient tool, but not a transformation of the mind, not a home for consciousness.

My verdict on the whole Super-AI topic and on transhumanism :  In history revolution was a term to characterize social progress and human pride. This "AI Revolution" is the most shabby inversion of the word revolution I can imagine. It is the most stuffy submission ever to the claim to power of others. (And 'others' are not machine gods but human slave holder  –NSA bureaucrats preparing the final war–  who have found their adoring idiots.)

'Transhumanism' is a term like 'Tausendjähriges Reich'. It has to obfuscate people's minds and to hide the true nature of a period in history. The 'Reich' lasted twelve years, 'Transhumanism' will have to last for presumably the next thirty years. After that time the real future (singularity in its worst meaning) will be accepted by the people in indifference and ignorance.


image via,  Los Alamos Maniac, IAS Princeton. (via Documentary by P. Calderon, 2014)

[1]  In the 1950s John von Neumann dressed his dreams of global dominance up and adapted his computational vision to the idea of weather simulation ('a little influence at an early stage leads to big changes later on'). – But of course he was interested in global prophecy rather - in the fate of the world and not in its weather.
Today's U.S. secret service is interested in this:  what happens if they could kill let's say Fidel Castro, Edward Snowden etc. before those could escape U.S. clutches, before those could become publicly known or do any liberating deed - if the secret service could execute preventive killings based on predictions by AI?

Von Neumann was the first who used the word singularity to describe this final stage of totalitarianism and slated it for mid-21st century. [*]

[*] Actually J. von Neumann (and later Ray Kurzweil) called it technological singularity. But they were wrong. I call it socio-singularity, it is an intended status of the society (intended by the master class to secure their rule).

Today Ray Kurzweil who sees himself as a successor in von Neumann's field is slightly dishonest on this point. He is lying (to say it more clear) and he deceives his audience. He says artificial intelligence is no more reserved for secret laboratories of a few governments in the world. These days everyone can be in control of artificial intelligence - of his smartphone or what ever. Kids in Africa have access to more informations than the president of the United States have had fifteen years ago..  and so on and so forth.

In fact Kurzweil knows full well he is turning reality on its head. Knowledge that could cause social or political change (a revolution) is more under hermetic (one- dimensional) control than ever before. A smartphone is a device not to control but to be controlled in most cases. Artificial intelligence we are talking about is not a matter of two hundred dollars but of one trillion dollars (5Eyes is spending on this). – So why does Kurzweil skew the truth?  Because he is just as much part of this terror of state as von Neumann was and because he shares the same vision.

The Arab Spring did not bear (doubtful) fruit because everyone has a smartphone and a Twitter account but because of the brigades of well-trained (and payed) fake contributors, downvoting bots and well-working blocking algorithms.  (Colorful balloons above the Tahrir Square at the right moment and the dirty job is done.)

[2]  The pre-charge punishment of Julian Assange and Norway's enmeshment in man hunt are not science-related topics. But they show the Swedish and Norwegian ties to the U.S. version of truth and policy. – In the Snowden case free and democratic positions in Norway asserted oneself.

[3]  I wouldn't be surprised if Bostrom, Tegmark, Goobar & co. are on the CIA's payroll under the item Infowars/ Disinformation.

[4]  What makes the best of sci-fi literature great? - It reflects about mind and existence and uses the idea of AI only to get deeper in the mystery of what existence is. Today's trivial sci-fi works the other way round, it's issue is not existence (being human) but technical aspects of AI. It is about how close AI can get to the human mind, how hostile or friendly it would be.

[5]  The new Kirk can't even verify he is someone else or not. - Only the old Kirk can (could).

[6]  Proof:  Instead of beaming you could duplicate. - Ask similar twins in real life: it is always someone else.

[7]  Pseudoscience as it is pushed by major U.S. research organizations is flourishing not only in sociology and IT ('singularity') but even in economics.
William D. Nordhaus uses the idea of singularity to suggest that unlimited economic growth and prosperity is only a matter of superintelligence applied in management and is only a matter of speed of computation.

It's sad when a professor of economics doesn't know prosperity is about humans  –not about profits–  and worths are made by human work and not by banks ( –it's the speed of humans' life, not the speed of computation– ).
Even more sad is this: the US hegemony and exploitation of the entire world works even with such poor scientific foundations – the rest is done by intended obfuscation (they hazard the deadly consequences of unlimited growth in exploitation of earth's resources).

[8]  At this point ASI is just the same as the voodoo of another topic in physical science I criticised earlier: The Voodoo of the Bubble Universe.
Postscript 2020-06:
How long the voodoo of the Bubble Universe will last is unclear. But is seems, five years after my AI article the voodoo of AGI is over:
Ragnar Fjelland, Why artificial general intelligence will not be realized

[9]  Tim Urban's writing at his blog Wait But Why is in-depth and comprehensive writing. But it is also inherently affirmative writing (to a self-denying extent). Tim Urban circulates any wild or even absurd hypothesis of any pop science luminary. At the most he leaves a question unanswered, but you will not find analytic criticism, much less sharp contradiction or exposure of demagoguism.
I mention this ('Wait But Why' is not my subject) because Elon Musk invited Tim Urban to write about his (Musk's) big enterprises (Tesla, SpaceX etc.). – And that indeed is a subject in part II of my essay. Urban, of all people, is invited – that tells a lot about Elon Musk's intentions and the purpose of the OpenAI company.

[10]  Postscript 2019-05: Podcast about AGI and theses of David Deutsch: