, 47 min read
Consciousness and free will are illusions: you are just a robot
48 thoughts on “Consciousness and free will are illusions: you are just a robot”
, 47 min read
48 thoughts on “Consciousness and free will are illusions: you are just a robot”
Does that make consciousness and free will illusions though? It invalidates particular definitions of them as completely independent objects, but there’s something to consciousness even if it it’s an epiphenomena. If the universe is strictly deterministic, free will can still have meaning in terms of a system’s partial ability to predicts its actions, or as a particular self-reference system that interacts with decision making via assumptions of self-agency.
Intel chips right now have “predictors” that try to predict its future actions (they are called branch predictors). Also, the chip can tell you a lot about itself, including which operations it supports. Somehow, Intel has not felt the need to insist that its chips have consciousness… because it is not needed.
You just don’t introduce one notion after another, keeping it all… You only keep notions that are proving useful.
As far as I can tell, the concept of consciousness plays no useful purpose in science today.
Because the science isn’t reached the level when it could scientifically understand consciousness.
I don’t see how being a robot that is simply governed by physics invalidates the notion of free will or consciousness. Perhaps you need to define those things first?
Science does not work that way. You introduce notions as a way to explain things. So apples fall to the ground. Why? Because “gravity”.
Here, what I am saying, is that there is nothing that we observe that cannot be explained without having to introduce superfluous notions like consciousness and free will.
If someone insists that we have free will, then this person needs to flush out the theory and then provide means of testing it.
For example, I say that there is no such thing as “intelligent design” and that human beings are simply the result of evolution. When I say this, I do not have the burden of flushing out intelligent design.
If you think that the concept of consciousness arises out our study of nature, then you need to explain how…
I’m not sure I understand the context of the debate here. Kind of by their very nature, consciousness & free will seem impossible to prove (or disprove) empirically. Their existence is relevant from a philosophical standpoint, not a scientific one.
Implicit in the recent predictions regarding AI is the fact that computers may (or may not) acquire consciousness and free will.
Ah, but by your line of reasoning doesn’t materially change the debate. Simulating the function of a design we don’t even understand that we *perceive* as consciousness and free will is the same problem as trying to create it, no?
@Christopher Smith
I think it matters in the following sense. In the Terminator movie, that seems to have shaped so much of our view of AI… Skynet acquires sentience at a specific point in time. Before that point, the computer was just a computer. After that point, the computer has sentience, consciousness, free will… I think many people are waiting for this moment. This point in time when a computer becomes “sentient”.
Other people think that computers will never reach this state because there is “something more” to human beings… we are more than mere computers because we have “consciousness”.
But I don’t think that’s what computer science tells us. Our computers are already mathematically equivalent to us. A mouse is equivalent, in some sense, to a human computationally. An iPad is equivalent to both. Human beings have more computational power, and smarter software, than either the mouse or the iPad… but there is nothing really fundamentally different at a theoretical computer science level. There is no soul.
If what I describe sounds self-evident and trivial to you, then please go to your nearest philosophy department and say those things.
If we will use big complex systems requiring a few observers with different qualities working seemlessly as one entity with conflicts resolved on fly it can come handy…
the more humans interact with algorytms the less sentient they become
“If someone insists that we have free will, then this person needs to flush out the theory and then provide means of testing it.”
Someone who insists that we do not have free will also needs to flush out the theory and provide a means of testing it. The default in science is “I don’t know”.
At this time, there are to the best of my knowledge no theories of free will or consciousness that could be tested by observation. And that means that these two terms are not scientific ones for now.
Since I am sure that someone will bring up Occam’s razor, let me add that Occam’s razor is about comparing two explanations and preferring the simpler one. It’s not about preferring the simplest story about something that we do not understand yet.
Someone who insists that we do not have free will also needs to flush out the theory and provide a means of testing it. The default in science is “I don’t knowâ€. (…) At this time, there are to the best of my knowledge no theories of free will or consciousness that could be tested by observation. And that means that these two terms are not scientific ones for now.
Scientists do not go around saying “we don’t know whether there are fairies or unicorns so let us include them in our textbooks just in case”. No. What scientists say is that as far as we know there are no fairies or unicorns. Until proven otherwise, they don’t exist. The same applies to free will and consciousness.
It’s not quite the same. Consciousness and free will are concepts that philosophers have been discussing for centuries, because they are part of human self-experience. Fairies and unicorns belong to the realm of legends that hardly anybody, at any time in history, believed to be literally true. I suspect no scientists has ever seriously thought about unicorns.
On the other hand, many scientists have tried to integrate free will and consciousness into some theoretical framework. There has been some progress about consciousness, although the current scientific approaches deal only with some aspects of the philosophical concept. Free will is likely to remain forever outside of science, because the non-existence of free will is a basic assumption of the scientific notion of causality. You can’t have reproducible observations if some free-will-possessing agent can change things arbitrarily.
Consciousness and free will are concepts that philosophers have been discussing for centuries (…)
That philosophers discussed something does not make it credible. Philosophers spent a great deal of time discussing angels, or how matter is composed of 4 elements.
Free will is likely to remain forever outside of science (…)
Not unlike fairies and unicorns.
Umm…but isn’t the fact that this debate happening at all indicative of the fact that we have free will?And are you trying to tell me that when I opened this article my reaction was predetermined? . You don’t even know what consciousness is, how can you decide that free will doesn’t exist? So does that mean I’m merely an automaton whose every action is programmed into her? So does that mean that when I care for someone it’s predetermined?Well I’ll have you know that when I was in fifth and sixth I was the most selfish person ever, now I’m not.I care more about other people than I care about myself. And you still say that I don’t have free will? Don’t make me laugh.
And how do you know that your view is the right one?If you were just predetermined to come to it?
I can never tell that my view “is the right one”. But that’s no excuse to just believe whatever other people believe.
but isn’t the fact that this debate happening at all indicative of the fact that we have free will?
Software agents have debates all the time. All of distributed computing is based on the fact that we have different nodes with different point of views that must be somehow reconciled.
You don’t even know what consciousness is, how can you decide that free will doesn’t exist?
I don’t know what the Holy Spirit is exactly, how can I decide it does not exist?
I can decide it does not exist because I cannot observe it nor can I observe its effects scientifically.
The same is true of free will. There is no experiment that can measure free will. We have no reason to believe that brains can do things that computers cannot do, and we know that computers do not have free will.
Well I’ll have you know that when I was in fifth and sixth I was the most selfish person ever, now I’m not.I care more about other people than I care about myself. And you still say that I don’t have free will? Don’t make me laugh.
What makes you think that a computer program cannot have feelings?
And how do you know that your view is the right one?If you were just predetermined to come to it?
There is nothing “superfluous” about the concepts of consciousness and free will. They, like all other concepts, help us organize the percepts in our consciousness. And no that’s not a circular definition: I used the fixpoint function.
Also, to say that consciousness is an emergent property of computation does not deny its existence, it affirms it.
If they’re illusions, then who is being fooled? :^)
Seriously, I think that consciousness is pretty well supported as a thing, even if we don’t know much about how it comes to be. There are a wide range of injuries, illnesses, etc. that can cause people to exhibit deficits in aspects of consciousness.
If they’re illusions, then who is being fooled? :^)
You can fool a computer.
Seriously, I think that consciousness is pretty well supported as a thing, even if we don’t know much about how it comes to be. There are a wide range of injuries, illnesses, etc. that can cause people to exhibit deficits in aspects of consciousness.
Can you make this statement more precise?
> You can fool a computer.
Not one sophisticated enough to be able to simulate consciousness and free will. 😉
I believe in “consciousness” and “free will” as almost entirely compatible with determinism, but practically impossible to determine.
Your current cumulative mental state is “consciousness” with input as all your prior experience, sprinkled with a bit of quantum in-determinism. Largely (but not entirely) predictable if all the input were known, but in practical sense you do not know the input.
Your choices are “free will”, taken from your cumulative mental state and current experience. (With perhaps a hint of quantum in-determinism.)
The folk who want those boxes to be large, crude, and simple are wrong. The folk who get all mystical … not one of those. On the other hand the “quantum in-determinism” is a bit tricky. Might be nothing more than what we think we presently know. Might be a bit more.
Robots are largely devoted to meeting some chosen external desires. Humans are largely driven in-built evolutionary goals, but much differentiated my individual experience.
(Well … perhaps not so differentiated in all cases.)
Do we want robots to have that difference? Likely not.
Humans are not robots (though some states want this).
My (very) speculative wildcard is that I am not sure that experience and time is a one-way arrow. We assume so as a simplifying assumption, and that sort of assumption is usually right. Usually is not always. How to put this to measure is unclear, so this remains a simple speculation.
(Also later, as usual, wanting an edit button. Are personal weblogs dead on practical base?)
“As a computer scientist, it is natural for me to view the brain as a computer. ”
As an information processing device, yes, certainly. As a computer in the sense of today’s digital computers? I am not sure that’s a useful analogy. A brain is clearly more than a computer: at the very least, it is a computer with attached I/O devices. But my main point is something very different: I think we can learn more by looking at the differences in information between brains and computers than by looking at the similarities.
As an information processing device, yes, certainly. As a computer in the sense of today’s digital computers?
In computer science, “information processing device” and computer are interchangeable. Initially, when Turing started his work, a “computer” was actual a human being. There are plenty of computer scientists who work on quantum computing, DNA computing and so forth.
I think we can learn more by looking at the differences in information between brains and computers than by looking at the similarities.
A fundamental axiom of computer science is that there is no “information processing device” more general than a Turing machine. So unless you can demonstrate that the brain can do computation that a Turing machine cannot do, then they are mathematically equivalent. This is a very deep result.
There are plenty of differences between, say, a digital computer and a DNA computer… and these differences are interesting… but the foundational argument of computer science is that core computer science results apply to any computer, whatever its shape and form… and that includes the brain.
As a physicist, I was thinking of the differences in the implementation of computation rather than in the underlying definition of computation. I will happily accept, at least as a working hypothesis, that brains are Turing-complete but no more.
However, unlike digital computers, brains probably don’t implement computation as high-fidelity physical realizations of the mathematical models of computation. It looks like they use highly redundant noisy processes. That’s something that engineering has yet to explore as a viable technique.
Computer science has no problem with probabilistic algorithms. Computer science has also no trouble building reliable computers out of unreliable parts.
Could not consciousness simply be the observations of a computer observing the computer? Here is how I understand consciousness:
We have a cerebellum which could probably be equated to a computer such as a robotic system. Unlike most robots, it can also determine “threat” levels and instigate some reflexive responses.
Then we have frontal lobes (probably where consciousness arises). These lobes monitor, and can drive the cerebellum. They are “aware” of what happens in the cerebellum, and store historical information, by which they can “decide” on courses of action – which may be beyond “fight or flight” which the cerebellum is capable of independently.
This is not magical, it is just science. But it is still consciousness.
Could not consciousness simply be the observations of a computer observing the computer?
An Intel processor observes itself. For example, if it is being very active, it will increase its power usage and its frequency. If the demand is less, it will lower its frequency. Are Intel processors conscious?
How does it react to the new input it wasn’t designed to react to? or to partial loss of own capabilities?
Observer of higher level can derive knowledge from own past evolution as it has some muscle memory of past experiences, from “previous lives”, intuition and intelligence capacity built up jumping up on selforganisation competences ladder solving problems from one cell level to complex multicellular organism.
Processor is not evolving,adapting own functions, serving in a system of different organs with different levels of functionality and responding to changing conditions around it it is non separable part of.
If you learn about brain without considering whole body it evolved in you won’t ever understand all purposes and functions it’s built for and to need. This also includes environment this body was adapting to evolving.
I’m not sure why you say that “There is no need for action at a distance or mysterious quantum effects.” There are certainly plenty of “mysterious quantum effects” in the real world, so there is no reason that they could not be operating in the brain and mind as well. Likewise, it is not clear that “action at a distance” is a correct description of anything that actually happens, but there are certainly real things that resemble that, and there is no reason why these things could not be involved in the brain and mind as well.
There are certainly plenty of “mysterious quantum effects†in the real world, so there is no reason that they could not be operating in the brain and mind as well.
There is actually a pretty good reason: scale. Neurons are large cells, and the brain is itself quite big compared to, say, an Intel processor.
There is also the fact that, well, we have no testable scientific theory of the mind that would involve “mysterious quantum effects”. It is all pre-scientific wishful thinking.
So what is the implication if the brain proves impossible to model? Would that suggest some metaphysical process is in play beyond computation?
So what is the implication if the brain proves impossible to model? (…) Would that suggest some metaphysical process is in play beyond computation?
There are many things in biology that we do not understand. We don’t know what aging is. We don’t know where the biological clock is. We don’t know why we get cardiovascular diseases. We don’t understand Alzheimer’s. Even cancer remains something of an enigma. DNA is still gibberish to us.
The fact that we do not understand something does not justify invoking hypothetical entities like a soul, a consciousness or magical unicorns.
Nobody says “oh… we don’t understand how DNA can turn a cell into a lung, so there must be magical quantum effects”. Why do we insist that the brain must be somehow subject to magic?
CPU is a processing unit. Program feeds a set of instructions into it. It seems our brain is hosting a number of programs from the vey basic to complex. Consciouness could be just another program that tends to run in background and examine execution logs, putting a spin on particular experience. Like everything in our body, it must serve a purpose; perhaps ranking experiences. It does mean that turning a particular program off makes us less human.
I like to think of free will as a species implementation of pseudo-randomness. It’s an algorithm based on impartial knowledge. It shares the same characteristic with software random gens; predictable, periodic, fairly repeatable.
Like everything in our body, it must serve a purpose (…)
But even if I were to agree that everything in the body must serve a purpose… is consciousness in our body? Where is it?
In science, there are different levels of description and abstraction.
Also, science has explanations for some optical illusions such as mirages.
When someone claims that scientifically speaking free will and consciousness are illusions, then even if we accept this conjecture we could still ask if there is a good explanation of why many ‘see’ the illusion of free will and consciousness.
As we learn more about how the brain works, we’ll establish a better understanding of the materialist claim that the mind is what the brain does. But as our understanding of this develops, it may well involve different levels of description and abstraction.
The question of which is best level of description is determined by the problem at hand. To describe the workings of a computer, for example, you could pick from the following: logic gates, machine language, assembly language, a high level programming language (e.g. Python), or the user interface.
As for the discussion of free will and consciousness in the human brain, you can pick from the following levels:
– an individual neuron,
– a group of neurons (btw: Kurzweil claims that it takes an average of 100 neuron in the cerebral cortex to form a pattern recognizer),
– groups of pattern recognizers (Kurzweil claims there are around 300 million pattern recognizers in the cerebral cortex), or
– one’s 1st person account of one’s stream of consciousness (for lack of a better term), which for many of us comes in the form of a linear sequence of words that we hear only with our mind’s ear.
It may well turn out that we would not want to apply the terms ‘free will’ and ‘consciousness’ to the behaviors of individual neurons.
The behaviors of neurons and groupings of neurons may be explained as emergent
phenomena. We do this in many areas of science. For example, when we explain the phenomena of boiling water at an abstract level we do so without going into the details of trying to explain and predict the behavior of individual water molecules.
Of course, human brains are not the only kind of brain. Animal brains are different from human brains in many ways. So, we may want to develop better computer simulations of the brain (along with theories about how the illusion of free will and consciousness works) by reverse engineering brains of different sizes and of different levels of complexity.
We may also want to artificially evolve intelligent agents in virtual environments starting low level intelligent agents, with simple artificial neural nets, and then progressing to more complex agents with more complex neural nets.
With computers, we know how to map from one level of description to another (e.g. from logic gates to machine language to assembly language).
But, with biologically evolved brains, we still have a long way to go before we’ve completely reverse engineered the human brain. As we improve our ability to simulate the brain, we’ll learn a lot more about how what we now call the mind and consciousness emerges (or how the illusion of the mind emerges) from the behavior of the brain.
This comment can also be found on my blog at:
Level of description and abstraction for the brain, computer, and mind
http://innovationmemes.blogspot.com/2016/01/level-of-description-and-abstraction.html
(…) we could still ask if there is a good explanation of why many ‘see’ the illusion of free will and consciousness.
Absolutely.
Animal brains are different from human brains in many ways.
Certainly, we have bigger brains than mice do, but a mammalian brain is a mammalian brain.
To my knowledge, we have not yet figured out a mouse’s brain. I suspect that once we do, figuring out a human brain will be easy.
One more comment: We generally think of biological evolution as being creative rather than deterministic because evolution draws on the random when copying errors occur. Although we normally think of computation as being deterministic, it doesn’t have to be. Finally, we can think of evolution in a general sense as being a probabilistic computational process.
For more on this last idea, see:
Evolution as Computation
http://innovationmemes.blogspot.com/2015/12/evolution-as-computation.html
We generally think of biological evolution as being creative rather than deterministic because evolution draws on the random when copying errors occur. Although we normally think of computation as being deterministic, it doesn’t have to be. Finally, we can think of evolution in a general sense as being a probabilistic computational process.
Evolution is definitively a computational process.
Without a plausible definition of consciousness, the discussion is a moving target. I tend to think of consciousness as a governor checking on my mental self and bringing awareness to rather autonomous processes inside. There could be a brain region(s) responsible for the function. That’s out of my league.
Without a plausible definition of consciousness, the discussion is a moving target. I tend to think of consciousness as a governor checking on my mental self and bringing awareness to rather autonomous processes inside. There could be a brain region(s) responsible for the function. That’s out of my league.
I am saying something very precise. I am saying that you are nothing more than a robot. Your brain is a (biological) computer. Nothing more. It is mathematically equivalent to a Turing machine.
So, I am saying that if you claim to have free will and consciousness, then so does your computer.
Now. If you want to say that computers have consciousness and free will… then we are in agreement that this becomes a moving target… but, largely, people don’t think of their laptops as having free will.
If however, you are saying that you have free will and consciousness, while your laptop doesn’t… then we are clearly in disagree. No moving target.
What a strange argument. Your responses strike me as those of someone who has never engaged the basic philosophical questions of cognitive science and philosophy of mind. It should be plainly obvious that the difference between a human mind and a laptop, while both are computers, is that the human mind has reached a level of complexity that gives rise to the phenomenon of qualia and/or consciousness (in the broad sense of a feeling of “self knowing”). Arguing that one can’t call humans conscious without also calling laptops conscious is patently absurd, as any expert in the subject will tell you. Free will may be an illusion, but consciousness is an actual phenomenon that every person experiences. Not knowing the nature of that experience does not mean it is an illusion. Comparing it to unicorns and fairies (which are the exact opposite in terms of evidence – no one experiences them, but everyone experiences consciousness) is equally absurd.
Your responses strike me as those of someone who has never engaged the basic philosophical questions (…)
Ad hominem. Please take aim at the arguments, not their author.
as any expert in the subject will tell you
That’s an appeal to authority and I put very little faith in such arguments.
Not knowing the nature of that experience does not mean it is an illusion. Comparing it to unicorns and fairies (which are the exact opposite in terms of evidence – no one experiences them, but everyone experiences consciousness) is equally absurd.
You’ll find millions, if not billions, of people who are convinced that God exists and that they see proof of his existence all around them. You will also find people, even today, who swear that they experience the presence of faeries.
People report experiencing many things on a routine basis that are not really there.
I already posted my opinions on this topic previously, and they pretty much in line with most of the other posters in favor of consciousness being real.
Most people will tell you they are conscious. Since I myself *know* from my own experience that I am conscious, I infer that others are their own sentient beings as well since they appear to be the same animal.
If we disagree on this point, then we can argue all day and never get anywhere.
Very interesting.
For a recent (mid 2017) professional view on free will attributed to not-necessarily human subjects, please see “Can a robot have free will?” which I wrote for the journal ‘Entropy’. Consciousness is not required for free will according to the definition I use for free-will (everything turns on definitions in philosophy).
I define free will as the condition in which all of the following are jointly true:
FW1: there exists a definite entity to which free-will may (or may not) be attributed;
FW2: there are viable alternative actions for the entity to select from;
FW3: it is not constrained in the exercising of two or more of the alternatives;
FW4: its `will’ is generated by non-random process internal to it;
FW5: in similar circumstances, it may act otherwise according to a different internally generated `will’.
Based on this, I find two primary conditions for free will in the deep sense:
1) that the agent be autonomous (this is defined by the condition of transitive closure of causal links among the agent’s parts) – roughly the Kantian whole;
2) that the agent creates its own internal model of itself and pertinent environment, with which it evaluates plausible actions against its own internally generated goal (maximising an objective function), selects one and acts upon it, in a cycle.
According to my analysis, this computational requirement could be achieved with a pair of Turing machines and a finite-state automata – the free-will machine, which is hypothetical, but represented in figure 2 of my paper.
I have not yet given the same treatment to consciousness, but it seems to me that consciousness is really a psychological term used to describe what it is like to be aware of oneself in a particular way (which we could in principle specify). It refers to a subjective experience and it is a matter for psychologists and neuroscientists to work out the physical-world (matter, information, ‘wet-ware’) that corresponds with, at least parts, of the experience of consciousness. I think that since it is a subjective human experience, we could only ever speak of it as an analogy in anything other than a human being.
There are two questions that commonly come up:
1) What will happen if I abandon the illusion of free will?
Nothing. You never had it in the first place.
2) How to make a robot think it has free will?
Just program it to think it has one. (just what has happened to us from culture, social interactions, language induced fallacies etc.,)
cheers
Hmm, let´s do the same but with some good ol´ negation:
1) What will happen if I abandon the illusion of not having free will? Nothing. You always had it. So there.