This is a fundamental and unquestioned assumption stemming from 19th century thinking and made fashionable by the era of computers. But is it actually valid?
An algorithm is purely deterministic and finite in nature. That is the very definition. If you understood this, you'd realise that algorithms cannot give rise to minds (consciousness), but only automation. Just as a machine cannot create energy out of no where (they only transform it), an algorithm cannot create information from nothing. In fact, they can only deterministically transform information from one form to another which -- if you think about it -- is all that computers do.
And just what the hell is "information" anyway? We say, "Oh yes information is facts and figures and statistics and such like," but it is rarely understood.
For example, is information a property like energy which can be neither be created or destroyed? Do algorithmic processes dissipate information, just as physical machines dissipate heat? Does a “particle of information” have a mass (some like Melvin Vopson argue that it does)? And can information actually exist without mind? Indeed, does knowledge require a knower? And what does it mean to "know" something?
Roger Penrose says that consciousness is the ability to break the rules, and argues in his late 80s book, “Emperor’s New Mind”, that the human brain is not algorithmic but has deep connections to quantum mechanical world. And the quantum mechanical level is where things get very weird when it comes "knowing things".
At its heart, the universe is not deterministic in nature. The implications are more than fascinating, as described below:
I believe that the idea of AI is used as a tool to provoke argument and agitation against anyone opposing the narrative of progressives left leaning liberals.
Thanks for the great article! I can’t say I have the technical background to add much evidence to one hypothesis over another when it comes to the dangers of AI, although I love your heuristic: global crisis + global solution + censorship = fraud. My ignorance makes AI the perfect topic for fraud, just as Covid did 3 years ago. With respect to intuition, I think non-general AI’s might still pose a danger (as well as tremendous benefit), so there may be degrees of potential challenges, with extreme cases being general AI taking over the world and AI never posing a threat whatsoever. With respect to institutional capture, I offer the entertaining bill gurley talk in case your audience has not seen it yet. https://www.youtube.com/watch?app=desktop&v=F9cO3-MLHOM&si=13_g_ZT3sT2JkoKJ
I fully concur that AI is not the threat it is made out to be, and that our ability to ever develop true AI is very likely a pipe dream. The use of chatGPT makes it clear it has been biased philosophically by its programmers, and so is not true AI.
I would challenge, however, a few aspects of the sentence: "So my intuition is that no simulation likely to bring about the evolution of general-purpose intelligence is plausible, even though we can say, with certainty, that it is possible, not least of all because we know it has happened once; and since no non-evolutionary development of a general intelligence is possible, I infer that there is no cause to fear a singularity."
Although we know with certainty that humans have general-purpose intelligence, we do not know with scientific certainty how it came about. To claim we know we obtained intelligence by evolution is incorrect. Evolution is an unproven theory, one with numerous and serious flaws and/or problems, and which distinctly lacks sound scientific proof. Therefore, to judge the potential for AI based on evolutionary notions is necessarily limiting in nature.
Furthermore, it is also incorrect to say that "no non-evolutionary development is possible." As it would be difficult to prove by scientific rigor precisely how mankind obtained its unique and miraculous intelligence, so also it is impossible to say with certainty how many possible paths there are to developing intelligence.
I for one am persuaded that our intelligence in and of itself disproves evolution, even as Nick here has argued that we will never be able to program intelligence into a computer. If man, (with all his intelligence directed to that singular end) is incapable of developing AI, surely random evolutionary forces are less likely to do the same in a living body! Living cells, immune systems, intestinal tracts, lungs, and so forth are masterpieces uniting physics, biology and chemistry, and so complex as to be beyond our grasp to this day, despite millions of scientists working endlessly on unraveling their mysteries. Human intelligence stands far and beyond relatively simple things such as lungs or immune systems. It defies all logic to imagine any viable process or path by which such complex and incomprehensible building blocks might be assembled on their own, let alone all together and simultaneously in one human by evolutionary forces. As many academics and researchers of all ages have concluded, human intelligence in a human body proclaims a creator and a gift.
I enjoy el gato malo’s take on AI... the reputation economy is already taking shape.
they do not fear that AI will lie to us.
they fear that AI will tell us the truth about just how much our “leaders” have lied to us.
AI can be the tool that defends us from the state or the tool the state uses to insert untold and unprecedented forms of suppression and dominance into the fundamental substrates of our lives.
Nov 23, 2023·edited Nov 23, 2023Liked by Nick Hudson
I agree with your conclusions, though for slightly different reasons. We ran into the differing semantics thing on TwiXter (constrained by character limits and all), but this follows on directly from what I was saying there.
If you take 'intelligence' to be the organising principle, the fundamental process operating within all living systems which brings them into an ordered dynamic equilibrium persisting through time and space out of entropic random chaos, and 'creativity' as the application of that intelligence through the vehicle of living systems, then both must necessarily be the drivers of evolution rather than emergent properties of it. (Some folks give that intelligence names like God or Allah or the Tao or whatever, but I think that risks distorting it. It's that which cannot be named.)
Both intelligence and creativity in this conception are universal: present in all living systems (rather than a reserved property of Homo sapiens as our hubristic imaginings would have us believe). Hence resulting in an almost infinite multitude of different experiences of life far beyond the very limited human realm of sensing and experience.
An illustration to put this in context ... Ask ChatGPT how long it would take the world's largest supercomputer to model the life processes of a single eukaryotic cell. It'll likely tell you it's extremely complex and that the time taken would depend on the level of detail and complexity of the model. Then it'll likely tell you that it's such a challenging task it may not be feasible with current technology. (Someone else I know who asked this question came back with an estimate of 390 million years.)
The fear of the 'singularity' (which always used to be the term for the centre of a black hole in my picture books) connects to something fundamental in us. I think it's some deep intrinsic 'knowing' that intelligence is somehow synonymous with life. Hence the worry that if we create something intelligent enough, it will come alive and batter us. I don't know if that's possible. It might be. But I do know that it makes the term 'artificial intelligence' oxymoronic. And that next to the Life intelligence that in-forms us (and which renders AI completely obsolete would but we realise it ...), the sum total of human knowledge (not intelligence) as aggregated by our clunking machinery is beyond risible. We're light years away from creating an intelligence capable of battering us by those means.
Meanwhile, as C G Jung so presciently remarked, it is "... man himself who is man's greatest danger to man, for the simple reason that there is no adequate protection against psychic epidemics which are infinitely more devastating than the worst of natural catastrophes."
"When we witness such behaviour are we seeing an algorithm manifest its own intentionality, or the manifestation of a programmer’s intentionality expressed through that algorithm? There’s no doubt it’s the latter."
It is also the programmer's intent expressed through the selection of data used to "train" the AI.
>Human minds are algorithms.
Stop right there!
This is a fundamental and unquestioned assumption stemming from 19th century thinking and made fashionable by the era of computers. But is it actually valid?
An algorithm is purely deterministic and finite in nature. That is the very definition. If you understood this, you'd realise that algorithms cannot give rise to minds (consciousness), but only automation. Just as a machine cannot create energy out of no where (they only transform it), an algorithm cannot create information from nothing. In fact, they can only deterministically transform information from one form to another which -- if you think about it -- is all that computers do.
And just what the hell is "information" anyway? We say, "Oh yes information is facts and figures and statistics and such like," but it is rarely understood.
For example, is information a property like energy which can be neither be created or destroyed? Do algorithmic processes dissipate information, just as physical machines dissipate heat? Does a “particle of information” have a mass (some like Melvin Vopson argue that it does)? And can information actually exist without mind? Indeed, does knowledge require a knower? And what does it mean to "know" something?
Roger Penrose says that consciousness is the ability to break the rules, and argues in his late 80s book, “Emperor’s New Mind”, that the human brain is not algorithmic but has deep connections to quantum mechanical world. And the quantum mechanical level is where things get very weird when it comes "knowing things".
At its heart, the universe is not deterministic in nature. The implications are more than fascinating, as described below:
https://informationphilosopher.com/freedom/laplaces_demon.html
You are correct in saying that AI is basically automation, but have not understood the reasons.
Moreover — and this is contentious I know — but I would argue that it is not atoms and quarks that are axiomatic in nature, but consciousness itself.
Thank you Nick for an enlightening article.
I believe that the idea of AI is used as a tool to provoke argument and agitation against anyone opposing the narrative of progressives left leaning liberals.
It's a tool for distraction.
Thanks for the great article! I can’t say I have the technical background to add much evidence to one hypothesis over another when it comes to the dangers of AI, although I love your heuristic: global crisis + global solution + censorship = fraud. My ignorance makes AI the perfect topic for fraud, just as Covid did 3 years ago. With respect to intuition, I think non-general AI’s might still pose a danger (as well as tremendous benefit), so there may be degrees of potential challenges, with extreme cases being general AI taking over the world and AI never posing a threat whatsoever. With respect to institutional capture, I offer the entertaining bill gurley talk in case your audience has not seen it yet. https://www.youtube.com/watch?app=desktop&v=F9cO3-MLHOM&si=13_g_ZT3sT2JkoKJ
I fully concur that AI is not the threat it is made out to be, and that our ability to ever develop true AI is very likely a pipe dream. The use of chatGPT makes it clear it has been biased philosophically by its programmers, and so is not true AI.
I would challenge, however, a few aspects of the sentence: "So my intuition is that no simulation likely to bring about the evolution of general-purpose intelligence is plausible, even though we can say, with certainty, that it is possible, not least of all because we know it has happened once; and since no non-evolutionary development of a general intelligence is possible, I infer that there is no cause to fear a singularity."
Although we know with certainty that humans have general-purpose intelligence, we do not know with scientific certainty how it came about. To claim we know we obtained intelligence by evolution is incorrect. Evolution is an unproven theory, one with numerous and serious flaws and/or problems, and which distinctly lacks sound scientific proof. Therefore, to judge the potential for AI based on evolutionary notions is necessarily limiting in nature.
Furthermore, it is also incorrect to say that "no non-evolutionary development is possible." As it would be difficult to prove by scientific rigor precisely how mankind obtained its unique and miraculous intelligence, so also it is impossible to say with certainty how many possible paths there are to developing intelligence.
I for one am persuaded that our intelligence in and of itself disproves evolution, even as Nick here has argued that we will never be able to program intelligence into a computer. If man, (with all his intelligence directed to that singular end) is incapable of developing AI, surely random evolutionary forces are less likely to do the same in a living body! Living cells, immune systems, intestinal tracts, lungs, and so forth are masterpieces uniting physics, biology and chemistry, and so complex as to be beyond our grasp to this day, despite millions of scientists working endlessly on unraveling their mysteries. Human intelligence stands far and beyond relatively simple things such as lungs or immune systems. It defies all logic to imagine any viable process or path by which such complex and incomprehensible building blocks might be assembled on their own, let alone all together and simultaneously in one human by evolutionary forces. As many academics and researchers of all ages have concluded, human intelligence in a human body proclaims a creator and a gift.
I enjoy el gato malo’s take on AI... the reputation economy is already taking shape.
they do not fear that AI will lie to us.
they fear that AI will tell us the truth about just how much our “leaders” have lied to us.
AI can be the tool that defends us from the state or the tool the state uses to insert untold and unprecedented forms of suppression and dominance into the fundamental substrates of our lives.
that’s the real choice.
https://open.substack.com/pub/boriquagato/p/ai-reboot-control-altman-delete?r=rd4dp&utm_medium=ios&utm_campaign=post
https://open.substack.com/pub/boriquagato/p/settling-in-to-the-reputation-economy?r=rd4dp&utm_medium=ios&utm_campaign=post
I agree with your conclusions, though for slightly different reasons. We ran into the differing semantics thing on TwiXter (constrained by character limits and all), but this follows on directly from what I was saying there.
If you take 'intelligence' to be the organising principle, the fundamental process operating within all living systems which brings them into an ordered dynamic equilibrium persisting through time and space out of entropic random chaos, and 'creativity' as the application of that intelligence through the vehicle of living systems, then both must necessarily be the drivers of evolution rather than emergent properties of it. (Some folks give that intelligence names like God or Allah or the Tao or whatever, but I think that risks distorting it. It's that which cannot be named.)
Both intelligence and creativity in this conception are universal: present in all living systems (rather than a reserved property of Homo sapiens as our hubristic imaginings would have us believe). Hence resulting in an almost infinite multitude of different experiences of life far beyond the very limited human realm of sensing and experience.
An illustration to put this in context ... Ask ChatGPT how long it would take the world's largest supercomputer to model the life processes of a single eukaryotic cell. It'll likely tell you it's extremely complex and that the time taken would depend on the level of detail and complexity of the model. Then it'll likely tell you that it's such a challenging task it may not be feasible with current technology. (Someone else I know who asked this question came back with an estimate of 390 million years.)
The fear of the 'singularity' (which always used to be the term for the centre of a black hole in my picture books) connects to something fundamental in us. I think it's some deep intrinsic 'knowing' that intelligence is somehow synonymous with life. Hence the worry that if we create something intelligent enough, it will come alive and batter us. I don't know if that's possible. It might be. But I do know that it makes the term 'artificial intelligence' oxymoronic. And that next to the Life intelligence that in-forms us (and which renders AI completely obsolete would but we realise it ...), the sum total of human knowledge (not intelligence) as aggregated by our clunking machinery is beyond risible. We're light years away from creating an intelligence capable of battering us by those means.
Meanwhile, as C G Jung so presciently remarked, it is "... man himself who is man's greatest danger to man, for the simple reason that there is no adequate protection against psychic epidemics which are infinitely more devastating than the worst of natural catastrophes."
Feels like Aaron Swartz open source deja vu moment.. perhaps a pinch of John Parry Barlow too!
https://www.eff.org/cyberspace-independence https://web.archive.org/web/20111120050620/https://www.1000manifestos.com/aaron-swartz-the-guerilla-open-access-manifesto/
"When we witness such behaviour are we seeing an algorithm manifest its own intentionality, or the manifestation of a programmer’s intentionality expressed through that algorithm? There’s no doubt it’s the latter."
It is also the programmer's intent expressed through the selection of data used to "train" the AI.
OPINION: "The simple Reason Why ISRAEL Invaded GAZA" - Dr Vernon Coleman .......................................................
https://rumble.com/v3xbf9l-nov-17-2023-why-israel-really-invaded-gaza-the-shocking-truth-behind-the-ge.html
Fascinating!