I think that ant example answers that question with a strong "yes".
Think of Conway's Game of Life, which has been shown to be turing-complete. The various "creatures" that exist in its worlds exist even without the program being run. We run the program only to observe them... the execution is for the external observer's benefit, the creatures exist in the "mathematical space" of the cellular automaton, not on the computer where the simulation is run. If an algorithm can be conscious, then so can one of this game's creatures, and its consciousness will be observing and interacting with the Game's mathematical reality, which will seem quite physical to that creature even without ever being simulated on any "physical" (to us) hardware.
Maybe all of existence is like that, no? There are people who say that maybe our Universe is a simulation. I say sure, but it doesn't need to be "running" on anything... it just exists because the rules it follows exist mathematically. It does get simulated, to an approximation however... inside our conscious observation of it! The Universe "exists" mathematically, but a subset of it "runs" in the brains of the observers to which it gives rise... it is the ultimate strange loop, forever eating its own tail.
You’ve put the idea more succinctly than I could have, so thanks for that.
Well, the ideality one project in its representation of an algorithm surely might feels like that. An execution of a software in perpetually moving execution stack in a broader network of extremely volatile context is an other affair.
Well, these are not just algorithms, they are actual agents. So they are embodied and embedded in the environment. Each action or movement can change the information that this algorithm learns from. So different previous experiences mean different agents. Also, the learning process is reliant on noise, and this can cause different outcomes. You would have to reproduce the whole environment in order to get to a situation that an agent will produce the same results given the same input.
Also, laws of physics are like a fixed algorithm our brains run on.
> So where is the consciousness, in the execution, or in the algorithm?
Consciousness is in the triad formed of environment, agent and reward signals. It's a continuous loop of perception, judgement and action, followed by observing the reward signals. The purpose of this loop, for biological agents, is self reproduction - so it is a self reliant ultimate purpose, it needs no external purpose except this one.
Is a human in deep freeze considered conscious?
Most of these ideas came from Jeff Hawkings thousand mind theory
In the second book, her brain is then duplicated across multiple colonies. Because the ant computer isn't as powerful as silicon based computers, one of her instances later realizes that she is a significantly compressed version of herself and that she doesn't have most of the old memories, capabilities or capacity for emotions that her human self had, presumably in large part also due to being transferred between three different substrates (flesh, silicon, ants).
[1] https://www.goodreads.com/book/show/25499718-children-of-tim...
The even stronger analogy between OP's article and the book is what happens before the spiders cultivate the ants, and the knowledge that the spiders take from them. Before the spiders were building ant-computers the collective ant colony was making its own conscious decisions and was actually winning the intelligence race for a while.
But the observation still doesn't answer whether either a group of humans, a digital neural network or an ant colony actually is conscious.
"conscious" is a language construct that defines an abstract concept that is only vaguely defined. Why? Because we haven't found a falsifiable theory that allows us to objectively state "x or y are conscious or not conscious"
One big hurdle is that if you assert that an ML algorithm is conscious, then you open the door to all kinds of implications that many humans don't like or disagree on. Such as whether we are unique and special, or that we may also be deterministic automatons and free will doesn't exist.
At that point, you get caught up in modern philosophical debates which have started with Kant and Hume.
That's only scratching the surface of the implications. The real question arises when we have to decide what moral rights consciousness proffers.
A lot of arguments for and against vegetarianism, for instance, focus on whether animals suffer in a way that we would understand as suffering. A lot of arguments for terminating a comatose life center around the lack of consciousness.
If you suddenly declare that a piece of software is conscious, you have to grapple with what it means for all these questions—and for that matter with what we do with the software itself. Is the software enslaved, having to perform the same thing over and over again? Does turning it off kill it? Does it want anything? Specifically, something other than what its lot is? Should it have a voice in deciding how it is used?
Should it get the vote?
Should agglomerations of humans? Say, corporations?
Should they get free speech?
Sci-fi books have explored a lot of these subjects, of course, but reality will be different, and we'll have to deal with it on its own terms.
But we could quickly settle this question if we checked some things first: does it need anything (have necessities)? does it learn? can it act on its environment? Can it evolve? Is it part of a group or alone? If no, then it's probably not conscious.
All the conscious agents I know have one and the same ultimate goal - to exist and reproduce themselves. All their other goals are just subgoals.
By my definition a virus and AlphaGo which was trained as a population of interacting agents with a winner survives rule are both conscious of their environment, which is just a board for AG.
If life was a cartoon maybe.
Although i'm sure many people would argue if it believes its conscious we have no right to dispute, as we cannot know and even may be essentially the same.
Ever threatened legal action against a multinational corporation?
I lean on defining the concept based on adaptation to the environment - consciousness is the function that adapts an agent to its environment. Its purpose is to safeguard the agent against external perturbations and achieve its own goals.
For example, how would we get food without consciousness. How would reproduction work? Consciousness has a vital role here. Evolution works at a slow pace, consciousness is required for quick adaptation, otherwise the penalty is death.
I think consciousness is being made into something transcendent, or unfalsifiable, or essentially different than physical processes because we like to make ourselves feel special in comparison to the world.
(Why should I care what happens to other people, and try to avoid harming them? Because they're clearly conscious, capable of joy, suffering, etc. What about other animals, like dogs and cows and chickens? That seems pretty obvious too, given our biological and behavioral similarities. What about molluscs? Hmm, there are some potentially important differences there. What about rocks? I can't know, but they have none of the features I usually associate with consciousness, and if they do have internal experience I have no idea what determines it, so I might as well continue to assume not.)
As soon as you define consciousness in functional terms, you make it tractable, but you also detach it from the thing we were originally wondering about. (The problem is that purely materialist or functionalist explanations always run into the question 'but why does there have to be internal subjective experience associated with these things/events/systems, and not others?'. Consciousness itself, in the 'qualia' sense, never plays a functional role in these explanations -- and if we weren't already assuming its existence, they would give us no hint that it exists.)
It's also fun to think of yourself as a giant city of smaller organisms - you've got white blood cell cops, you've got red blood cells which are basically cars on arterial highways, you've got factories and garbage trucks and libraries.
I never took any Chinese lesson. However, suppose I obtain a huge instruction manual that tells me exactly which sequence of Chinese characters I should use to reply to any sequence of Chinese characters you (a native Chinese speaker) give me. Do I "know/understand" Chinese?
Let me ask you another question - when I throw a ball at you your subconcious brain has to solve a differential equation to know how to use muscles to catch that ball.
Do you know how to catch a ball if you don't know the math but still can catch it?
Or the system knows how to translate from English to Chinese. Human language isn't simply about following translation rules, though. It's also about communication and expressing thought. Or participating in language games.
> Let me ask you another question - when I throw a ball at you your subconcious brain has to solve a differential equation to know how to use muscles to catch that ball.
Why suppose the neural network needs to solve differential equations? Is that the only way to learn to catch a ball?
Yes. You need to decide where to put your hand, how to orient it, etc in reaction to the ball movement.
The answer is a solution to the differential equations, and you cannot consistently get a good answer to an equation if you don't solve it.
The solution probably isn't symbolic but numeric, but that hardly changes anything - you still need a lot of math to consciously solve such an equation numerically.
> Or the system knows how to translate from English to Chinese. Human language isn't simply about following translation rules, though. It's also about communication and expressing thought. Or participating in language games.
human doesn't get to decide what to do in Chinese room experiment - (s)he is just a dumb CPU that does table lookups in a huge book. Every possible response to a sequence of previous messages is already written in that book (so it must be quite big ;) ).
You could completely automate it, remove the human and nothing would change from the outside.
Do you agree at least that a xor gate is doing math?
Right, but my point is that following a bunch of translation rules from one language to another is not the same thing as understanding a language. That's not what humans are doing when they use language.
> The solution probably isn't symbolic but numeric, but that hardly changes anything - you still need a lot of math to consciously solve such an equation numerically.
This is assuming the brain is doing math. Even deeper than that, it's assuming that math is something more than a specialized human language. That math exists in nature to be harnessed by neurons.
Also the Chinese room would express emotions, do word games etc, whatever's appropriate in the context. The people doing the lookups wouldn't know that they are writing a joke, but who cares about them?
> This is assuming the brain is doing math.
Well of course brain is doing math, see: 234-123=111 - this is math, my brain did this.
> Even deeper than that, it's assuming that math is something more than a specialized human language
If neurons arriving at solutions to math problems consciously is math, then why neurons arriving at these same solutions subconsciously isn't?
I posited this years ago and the conclusion is what we call "math" is the linguistic expression of what is already intuitive to us.
The other example I used was a mother cat will seek out a stray kitten...but how can she possibly count & keep track of the number of kittens she should have? The answers are at once obvious but deceptively difficult to put into words.
How big is the book?
Seems like a nitpick about a theoretical concession in a thought experiment, right? No, it's actually very important.
If the book contains simple instructions like if you receive character A then reply with B, etc. in a Choose-Your-Own-Adventure style, then it would have to be exponentially large. Too large to be able to carry out more than a few words of conversation, and certainly you would not be able to converse about math, e.g, you're asked what 一 plus 二 is and you say 三, etc. The book would rapidly become larger than the planet Earth with any reasonable conversation depth. This is not just practical problem.
Okay, so the book isn't a simple lookup table, fine. You'll have to have a piece of scrap paper and write down things to refer to them later. But once you do that, it's obvious that you've created a system of memory. It undermines the whole force of the thought experiment. The book was supposed to contain all the brains and you were just supposed to mechanically follow along without understanding what you're doing. But now you're doing complicated things like solving math problems and then converting the answers into Chinese. In order to give the book a finite size, we've given you a lot of work to do, and now it's totally reasonable to say that you do know how to write Chinese. You write Chinese by looking it up, the same way that real translators do!
The Chinese room thought experiment is much discussed but it's a pretty poor thought experiment. It handwaves away all the important parts of language in order to make an inscrutable point about machine intelligence. It neither sheds light on machine intelligence nor language.
> that tells me exactly which sequence of Chinese characters I should use to reply to any sequence of Chinese characters you give me
Then the question becomes - can a language model understand the meaning of the text it generates? Or does it assign its own meaning to the data?
I'd argue that there's a major difference between consciousness and self awareness or sentience.