One thing that can neither be proven nor disproven is that there exists an external mind-independent world. Why does it appear (according to our best science) that we live in a purely physical world devoid of qualities? Erwin Schrödinger–Mind and Matter
I don’t believe we’ll ever have AI without consciousness. If we are simply opening a new portal to consciousness (like having a baby) it will take a mathematical formula that is self examining.
Voluminous computations will not produce a conscious entity. AI would need consciousness to become intelligent. It is quite likely there is only one consciousness and many apertures, many openings into it—Our brains are receivers. If somehow we created a new form of consciousness it would be immiscible and likely catastrophic to it or us.
Just as everything is made of one process (the collapsing wave function) of a single, fundamental phenomenon, adding another form could be catastrophic. Since there is no evidence for a multiple consciousness model vs a single source model, I would think science should err on the side of caution —but I doubt they will.
Who could we trust to manage the source code?
28 thoughts on “AI vs Consciousness”
I guess it depends on how you define AI.
When I started in IT in the early 80s AI referred exclusively to what’s now called ‘AGI (artificial general intelligence)’ – machine capacities that fully simulate – or exceed – human intelligence.
Something like the Deep Blue chess playing program or Tesla’s homicidal auto drive feature would have been called a ‘expert systems’ back then. But marketing departments have seen fit to appropriate and broaden the notion of artificial intelligence to the point where a fridge that orders groceries over the internet is now considered ‘intelligent’.
But assuming AGI is possible and testable at all (a big assumption) I wouldn’t be inclined to deny its intelligence simply because it showed no signs of consciousness. We can’t even be sure other people are conscious, but it seems a little unfair to deny them intelligence on that basis (despite the efforts of many public figures to convince me otherwise). As David Chalmers points out with his ‘p-zombie’ thought experiments, it’s perfectly possible to imagine something that looks and acts exactly like a human that lacks consciousness. And Daniel Dennett insists we are all like that and we only imagine we’re conscious (physicalists like him have to engage in some pretty extreme rhetorical gymnastics to avoid admitting their world view is intellectually bankrupt).
But as ‘consciousness’ and ‘intelligence’ both lack consensus definitions I can’t see how to avoid any discussion of your viewpoint quickly descending into semantics.
I’m a bit surprised by the Schrodinger quote at the top. I’d have thought he had a better grasp of the history of science than that.
The reason “it appear(s) (according to our best science) that we live in a purely physical world devoid of qualities” is because when Galileo invented science-as-we-know-it he specifically excluded qualities in favour of quantities, so that ‘science’ would become the branch of natural philosophy which could be exhaustively defined and communicated in mathematics. The old dualist would have been surprised (and probably amused) at anyone who thought science could define all of reality. He was building a tool, not a metaphysics.
The problems arose when the Enlightenment swept away the credibility and relevance of non-scientific methods of examining or describing reality. That left a lot of newly minted secularists unmoored (or unhinged) from the certainty their insecurities demanded that had previously been provided by their gods and scriptures. So they turned to ‘Science’ as the new god-of-all-things and in doing so had to deny or blind themselves to the huge scope of reality that lies outside the remit of science, including all forms of subjectivity such as consciousness. Hence physicalism and scientism.
The Enlightenment didn’t succeed in killing our gods. It just hugely impoverished them.
LikeLiked by 3 people
His assumption was ultimately a non-dual conscious model.
“Schrödinger assumed that there really is only one mind. Schrödinger thought this is supported by “the empirical fact that consciousness is never experienced in the plural, only in the singular. Not only has none of us ever experienced more than one consciousness, but there is also no trace of circumstantial evidence of this ever happening anywhere in the world.”
What you call the “physical world” is merely a highly-simplified representation of non-dual consciousness. We only see a piece of it making us feel isolated, but that is illusory.
It does depend on how one defines it. If a machine could act conscious why not us? A body can be in or out of consciousness. Maybe the terms should be in-conscious and un-conscious.
Those that claim the experience of cosmic consciousness certainly present a pretty uniform description that is compelling.
You might be the person to as (as a side note). Why do NDEs not mirror the experience of enlightenment or awaking in that sense?
You sure of that?
Setting aside the implicit assumption that either a body manifests consciousness or consciousness is manifested within a body, how would you know it can be in or out of consciousness?
There’s stronger evidence narrative memory formation depends on the state of a body. I can go out partying and not remember anything after I had, how many?, drinks. But, embarrassingly, my fellow party goers will insist I continued to be conscious for some time after than, then go into excruciating details about the manner of my consciousness.
I’ve also experienced retrograde amnesia preceding head traumas or anesthetic. Was I unconscious in the minutes prior to being hit or injected, despite continuing to behave just as idiotically as I usually do?
So I’m agnostic as to whether I’ve ever been unconscious or just unable to form memories. In fact, for any meanings of “I” I can think of the statement “I was unconscious” seems oxymoronic.
But it’s not just Daniel Dennett who claims to have freed himself from consciousness.
LikeLiked by 4 people
I don’t think Dan Dennet could handle the isolation of being disconnected from the grid. Or he is god.
Have a drink 🥃
And why does it surprise you Schrödinger would say “devoid of qualities”? Even today most people think Newtonian even though intellectually we live in a quantum world. Now the doom of space time is on the horizon, maybe people will get the quantum, that space time is only fundamental to this headset? There is quite a lag time.
I’m surprised he’d ask the question – unless it’s meant rhetorically. I’d have thought he knew quite well that the universe of science is devoid of qualities because they were cast out by Galileo when he created it.
According to Advaita, pure consciousness (chaitanya) is devoid of qualities (nirguna), but I don’t imagine Schrödinger would have equated chaitanya with the appearance of the world (maya).
He did not endorse traditional Western views that go under the names of reductive materialism and subjective idealism, but he found inspiration in non-Western, particularly Indian philosophies. As well as some other key players in the golden age of physics.
Yeah, I’m sorta familiar with the Upanishadic leanings of some of the early quantum physicists (though they’re exaggerated quite a bit by quantum mysticism gurus like Gary Zukav and Frijof Capra), but in your quote he’s not talking Advaita Vedanta, he’s interrogating a foundation of science.
He can lay off the waterboarding and third degree. Science ‘fessed up from the get go. It might take a bit of enhanced interrogation to get Daniel Dennett or tildeb to admit it though.
Science isn’t about qualities because it’s focused on quantities.
It’s blind to subjective experience because it strives for objectivity.
It’s not non-dualist because it’s reductionist.
Those things are its strengths. They only become weaknesses in the hands of those who seek to make science into something it’s not.
Science is a superb hammer, but not everything is a nail.
LikeLiked by 1 person
What are your thoughts on the doom of spacetime, Nima Arkani-Hamed, Ed Whitten, and Sean Carrol? Space time is not a fundamental property and likely is just the playground of our headsets.
Well, I’m not sure if the incoherence is Arkani-Hamed’s or comes from Heelan’s editorialising, but if he’s trying to say quantum theory and our current formulation of space-time have limited shelf-lives I can only agree – as I would if he’d said the same thing about any scientific theory.
And it seems to me they’re both already looking a bit spotty, but not as stale and moldy as Arkani-Hamed’s string theory. At least quantum theory and space-time worked for a while, unlike the mess of silly string some cosmologists have been tangled in for decades. And gravity itself looks weaker every day as the ‘dark matter’ that’s meant to save it shows no more signs of turning up than another much heralded Saviour.
That said, I’m in no way qualified to take a guess at how long it will be until quantum physics and space-time stop delivering the goods and get replaced by something with more mileage.
It’s important to keep in mind that science isn’t about discovering the ultimate truth of the cosmos. It’s about stretching our perspectives so we can do more with the tiny, distorted fragment of reality our minds are capable of grasping towards. As such all science is doomed and always will be; by more effective science. At least until the human race meets its doom.
I suspect quantum physics, space-time and humanity will all reach their doom at about the same time. Civilisation probably hasn’t got long enough to need any new cosmologies.
LikeLiked by 1 person
Part of the incoherence is the discovery shakes the foundation of everything. QM and relatively is not adequate to explain consciousness because it is not fundamental. Finding a physical link to generate it has been completely fruitless in spite of many great minds working on it. And from the equations it’s looks like the answer is outside space and time. Everything we know has assumed it is.
Well, I wouldn’t call Arkani-Hamed’s musings a ‘discovery’ nor quantum physics a ‘foundation of everything’.
It’s been known for well over half a century that quantum and relativity physics are incompatible, so it’s certain at least one of them is ‘wrong’. But even considering the formulae we use to codify and communicate our discoveries as fundamental laws of nature – or even ‘foundations’ of science – is pretty messed up if you ask me. As with any language, mathematics speaks of only that of which it speaks. To imagine it’s saying everything about anything is akin to mistaking a signpost for a destination.
But when I said ‘incoherence’ I was talking about the conflicts and contradictions getting tossed around to try to simultaneously make the case that physics ‘knows’ almost everything and that it ‘knows’ almost nothing. Pop neuroscientists often tie themselves into those sorts of knots but I don’t recall seeing a reputable physicist do it before.
So, for example, we’re expected to believe both “we have studied fundamental physics for 2,000 years and understand the field very well” and “we are too early in our approach to re-thinking physics to believe that what we already know is the real thing”. (BTW, not sure where that 2,000 year figure comes from. 2,700 years ago Thales of Miletus was already laying down principles of physics we still use today while, OTOH, Galileo, Newton and Einstein all ripped up so many of the ancient verities that quite a few people think contemporary physics started with one of them.)
Despite the token nod to epistemological humility in the previous quote this account of Arkani-Hamed’s thinking reeks of the sort of hubris that plagued late 19th century physicists who thought their discipline had almost completely codified the laws of the universe, with just a couple of problems left to mop up. He seems to imagine we can come up with a grand unified theory by ditching fundamental planks of physics – as Einstein did with Euclidean geometry – without shaking up the whole structure so much that we no longer “have enough data to know that certain results are certain”. That sure isn’t how it panned out the last time physics underwent a paradigm shift.
LikeLiked by 1 person
Love your last two paragraphs and last line related to science. Sums things up quite well, IMO.
I think if you looked into the names I mentioned, plus Donald Hoffman, you’d find abject humility in their approach. Very open about the lack of certainty, but eager to develop a new approach to reality and technology that eliminates space time which favors a more environmental approach
Not fuckin’ Microsoft, that’s for sure.
Unless we want the entire phenomenological universe to turn blue and die at the worst possible moment.
LikeLiked by 2 people
I’m not so sure. I think the trick is simply ‘writing in’ a reward system. Exploration (intelligence) then has a goal to move towards
LikeLiked by 2 people
That sounds a bit like how behaviorists disable intelligence, by over-ruling it with conditioned response. Besides, any ‘intelligence’ so generated is going to be limited by the intellect of whoever (probably a committee) comes up with the ‘goal’ and the reward system for nudging towards it. There’s plenty of dystopian science fiction about machine intellects that fixate on a goal while lacking anything approximating human empathy or compassion.
Methinks any goal simplistic enough to be effectively motivated by predefined reward systems is unlikely to generate anything we’d recognise as intelligence, at least on scales smaller than the evolutionary ones that produced our sorry excuse for it.
I’m sure you’ve noticed how the ‘nudge’ rewards favoured by neo-liberal ‘meritocracy’ have resulted in dumbed-down public discourse and loss of faith in almost all elite run institutions, from government to churches to science & academia to the media to justice systems to corporations.
In fact the only institution that’s seen an increase in public trust (in the US at least) over the past 20 years is the military, despite its long run of fuck-ups. As we all know, ‘military intelligence’ is an oxymoron, but you can bet there’s more money going into the development of military AI than any other sorts.
BTW, here’s a pic of a street sign deliberately hacked to confuse Tesla’s machine learning auto drive feature.
The AI reads this as a ’45 mph’ speed limit sign. As you can see, even by Elon Musk’s modest standards AI still has a very long way to go.
LikeLiked by 1 person
I wonder if your continued cynicism would disqualify you from any new approach? I’m going to see how Hoffman, Whitten, Hamed, and David Gross tackle this. Nothing else scientific has panned out. Emergence has gathered no evidence in spite of a tremendous amount of work and wording. At least people are talking again. Maybe you have an approach worthy of discussion? I think it will take a beginners mind.
If you’re talking a new approach to the development of AGI then I’m afraid my cupboard is bare. I don’t think we understand the evolutionary emergence of organic mind enough to simulate it convincingly on a machine or even if it’s possible.
As is so often the case we’ve gotten caught up in our own metaphors and over-extended them, with our hubris coming along for the ride.
Once it was envisaging godhead as a sort of superhuman father figure then tacking all the emotional foibles and authoritarian tendencies that go with that onto our imagined graven image. Now we’ve taken the metaphor of the brain as a computer and neurons as logic switches and imagined we can therefore build a silicon mind in a box – or worse, upload our own consciousnesses into a box – without the need (or capacity) to define ‘mind’ or ‘consciousness’ in a manner that might give us a clue as to what we’re really trying to achieve.
I can see no way forward on this, but that’s OK as I can see no reason to want a way forward. Lets leave AGI in the science fiction movies where it belongs.
If you’re talking breaking the expensive logjam in theoretical and experimental physics I’ve got some rough sketch ideas though.
Arkani-Hamed seems to think that if we imitate Einstein’s paradigm shift we can achieve one of our own, though for some reason he also thinks it can be achieved without invalidating huge slabs of our existing physics. A member of his audience was even so pedestrian as to suggest we need to rethink geometry again because, you know, it worked for Einstein.
Obviously you don’t attain Kuhnian paradigm shifts by repeating what’s already been done, but without one I’d suggest academic physics is what’s doomed. We’re not spending ever increasing research budgets on ever diminishing returns because we’ve nearly run out of physics to discover. It’s because we’ve already picked the low hanging fruit of our current paradigm and started bouncing off the walls of our mathematical and conceptual models. And the next shift won’t enable the unification of our existing theorems into a final perfect synthesis. It will open up the field by asking a lot more questions than it answers.
Rather than geometry we need to start questioning some of the more fundamental components of our current science. I can think of two obvious candidates.
Godel showed us that set theory can never offer a complete description of anything. Set theory is fundamental to the practical application of mathematics. Numbers have no real world equivalents. To have two or more of something requires us to impose the sort of abstract essentialism that enables us to, for example, take completely unique objects and reduce them to simplified common elements sufficiently to be able to say “I have two apples”, instead of “I have one unique thing and another unique thing”.
If we can find a symbology that respects the simultaneous uniqueness and unity of everything we might be able to transcend the math based science Galileo bequeathed us and start codifying and communicating our knowledge in a way that’s both more specific and more general. I suspect that would show us another way forward.
Another line of attack is inductive logic.
David Hume pointed out almost three centuries ago that inductive logic – and therefore causality – can’t be validated. We know the sun will rise tomorrow only because it has always risen ‘tomorrow’. Some(non-)day that’s gonna be a wrong assumption. There’s no reason to believe any other scientific theories are any different. That’s why scientists bring so much grief upon themselves trying to disentangle correlation from causation. It’s also why Karl Popper said the test of whether a theory is scientific is whether it can be disproved; because no matter how many times you replicate experimental results you can never completely prove it.
I think the very notion of causality – for all its practicality – is wrong. I think it’s an artifact of how we perceive time and that’s almost certainly wrong too.
When Einstein threw out Euclidean geometry he forced us to reassess our ideas of both space & time and matter & energy as separate things. Maybe replacing causality with something less linear, more relational and less time-bound will give us the new perspective we need. At least it should dispense with all those pesky acausal paradoxes quantum physics keeps presenting us with. I’m agnostic as to whether it would enable us to resolve space-time, matter-energy and gravity into one overarching general theory. I suspect it would make the question redundant.
They have offered imo, a more reasonable approach—not to throw away space time and field theory, but projecting back into space time we would be able measure and verify based on the laws we have, that actually work in space time. In that regard space time is a useful interface, but it isn’t fundamental. It’s simply how things work in the illusionary state we are in. This of course is an attempt to solve the hard problem and little else. I l think it is a facilitating idea.
I see nothing in that Arkani-Hamed stuff you link to that even suggests he’s trying to come to terms with consciousness, much less the hard problem.
Unless I missed something he’s just another plodder trying finish what he imagines to be the unification of theories Einstein started. He’s wearing his loyalties on his sleeve by questioning quantum physics and space-time while refusing to consider that, say, relativistic gravity might be flawed or incomplete, despite the fact that the continued failure to detect dark matter suggests exactly that. He also seems unfazed by the way half a century of string theory has led to a proliferation of ever more complex mathematical constructs without delivering a single testable prediction or experimental framework. It’s the final divorce of theoretical math-based physics from anything that occurs in the observable cosmos.
Arkani-Hamed can see the same old furrows in front of his nose that theoretical physicists have been plowing to no avail for two generations and no further. An Albert Einstein he sure ain’t.
You’ll have to look a little deeper than the quick link I sent you. His colleagues agree that at 10*-33 cm and 10*-43 seconds space and time are irrelevant—and where it arises may very well be consciousness.
The evidence of black holes was in Einstiens math yet he missed it until Karl Schwarzschild. Even then Einstien was wont to admit it. This too is in the math.
BTW, I should point out there’s no reason to question the ‘reality’ of quantum theory as it makes no such claims.
The interpretation of quantum physics that’s held sway since the early 50s – the Copenhagen Interpretation usually attributed to Bohr, Schrodinger, Heisenberg and Oppenheimer – is anti-realist. It makes no claim to be describing how the universe is. It just offers a practical way forward in predicting and manipulating certain phenomena (e.g. the electronics that are enabling you to read this right now).
Copenhagen has waxed and waned in popularity over the decades with its most vocal critics being American physicists who have no time for the philosophy of science (they caricaturise it by suggesting it says nothing about physics beyond “Shut up and calculate”). But no other interpretation has exceeded its level of acceptance and it continues to be valid long after many of its rivals that insist on at least local realism have been disproved experimentally.
So I think it’s safe to say Arkani-Hamed’s claim quantum physics is ‘doomed’ because it doesn’t fully describe reality is wrong-headed. It never claimed it did or could. It just gets what it does done.
The reason it is doomed isn’t because of the things you’re stating. Assuming as a Hindu I’d think you’d appreciate that consciousness is fundamental, and it won’t be discovered through Qp
“I don’t believe we’ll ever have AI without consciousness.”
Hal 9000 would beg to differ. 😉
There’s more than ample serious discussion here, I just had to. Sometimes I just can’t help myself.
Hows it going Jim? 🙂
LikeLiked by 3 people
Good boss! Hal was it. How easy we forget. Should be easy enough to add a new portal into consciousness, but create a new one?
Lucky to get connected to you your website. Contains lots of knowledge.
Will read these posts carefully.
Thank you so much for bringing such thoughtful insights.
LikeLiked by 2 people
I see your views and somewhere agree too Jim. Debatable as already had been but we surely are going through a defining time.