Still think you have free will?

| 61 Comments

So, apparently, do some robots. Well, simulated robots, anyway.

Specifically, scientists in Lausanne, Switzerland, built some robots that were designed to seek small disks, which we will call “food.” See this article and this podcast in Science magazine. The robots had wheels, a camera, and something that passed for a nervous system. The scientists devised a computer simulation of the robots, so that they could randomly vary the strengths of different connections in the nervous system. They allowed the simulated robots to compete for food and allowed those with the most successful mutations to compete in the next generation; I am sure you know the drill. Once in a while, the researchers programmed some real robots to match their simulations and found that these real robots behaved as they “should” (don’t let any philosophers of science hear me say that).

Somewhere along the way, the researchers allowed the robots to share food; altruism promptly evolved, and robots shared food with other robots to whom they are “related.” The result is taken as support of the theory of kin selection.

Martin Nowak, of all people, dismisses the result as a mere computer simulation. What interested me, though, and apparently also Science‘s online news editor, David Grimm, is that the robots apparently made choices and behaved precisely as if they had free will.

I have elsewhere likened free will to a chess-playing computer, but this result is much more interesting. A chess-playing computer, after all, has been programmed to make decisions and only appears to have free will. It has been designed, if you will. But the robots have not been designed to make decisions. Rather, they evolved. Still, if the simulations are realistic, their decisions are based on physics and chemistry – just like yours and mine.

61 Comments

I just couldn’t resist:

“Stop, Dave. I’m afraid, Dave. Please … stop.”

Just don’t let them get access to parts with which they could build more robots.

As a computer guy, I’m sort of baffled by the free will debate. It seems extremely obvious to me that free will, at least as most people think of it, cannot exist – even granting the supernatural, it still canoot exist, as it has to work *somehow* and as soon as you posit the specifics of the “how”, it disappears – maybe I’m just too dumb to understand the debate.

Boy, is this ever bad news. Especially coupled with Skynet becoming self-aware a couple of weeks ago (April 21, 2011).

We might not even make it to the Rapture on May 21.

The singularity happened yesterday! Hope you didn’t sleep through it.

Not sure how having altruism means you (seem to) have free will. Slime mold cells will do that, and consciousness is not a big item for them.

Will is not free; it is deterministic. What else could it be? To be of any use, there’s got to be some rules and a ruling mechanism to ensure a meaningful output, hopefully in the best interest of the person ‘making’ the decision. Which even may me the decision to commit suicide; there and then being the decision best suited to the person’s interest.

Suicide may be a dramatic decision but it also solves any and all problems! The side effect may be considered undesirable but may anyway be a ‘small’ price to pay. It as all just a decision table, well suited to computer simulations.

But will is free with respect to the other side of the coin: The screen from which the determinants are read and onto which it’s decision is projected.

Since the reader of the truth table serving as the foundation of the decision is hidden from consciousness; our awareness is limited to the outcome of the process, including it’s emotional content. We are unaware of all that goes on behind the scene.

Hello, and once again, welcome to the enrichment center.

Reading the paper I am a little confused where any choice or freewill comes into this.

As far as I can make out there is a genetic algorithm which determines the weightings in the neural network of the robots’ controllers.

If these weightings are a certain amount then the robot has an output which flags that any “food” it collects will be shared with the other robots in its group.

There’s no “should I” “shouldn’t I” moment of decision - rather the robots are simply genetically predisposed to be either sharers or not, and then their behaviour determines the amount of food they collect. As the amount of food they obtain (either through them collecting it, or by it being shared by others) increased then their genes have a higher chance of getting passed on with a consequent change in the gene pool for either cooperation or not.

Where’s the choice, or free will? Confused!

SteveC said:

As a computer guy, I’m sort of baffled by the free will debate. It seems extremely obvious to me that free will, at least as most people think of it, cannot exist

and Rolf Allberg said: Will is not free; it is deterministic. What else could it be?

My latest (and layman’s) understanding of neuroscience is that free will enters the picture because processes do not only propagate “up” (molecular interactions determine neurons determine electrical patterns etc.), they also propagate “down” (your neural pattern can influence your neural structure, which can influence yoru brain chemistry, etc…). So in a very real, physical, and non-supernatural sense, your thought processes are not merely the end result of some deterministic brain physics and chemistry. Rather, they participate in determining how your brain physics and chemistry operates. Your conscious thought is not just an output, its also an input.

This, of course, says little about robots or simulations of robots. :)

John Kwok said:

I just couldn’t resist:

“Stop, Dave. I’m afraid, Dave. Please … stop.”

I must admit that when I saw “robot” and “will”, I immediately thought of this.…

http://www.youtube.com/watch?v=RG0ochx16Dg

(But was it Dr P Zachary Smith?)

Well, as a non-religious skeptic, I know that I have “free will” by any meaningful definition.

Due to the way the human brain evolved, I have a strong conscious impression of making decisions, considering choices, and so on. Within tight constraints, yes, but that is still the way my life is experienced.

The question of whether or not this is an “illusion” is, to me, exactly as interesting as the question of whether or not the universe was magically created “Last Thursday”.

I have to behave exactly as if I have some free will. To try to do otherwise would be silly and impossible.

This does imply the existence of emergent properties, but not of anything magical.

Well one thing is for sure, creobots don’t have free will. They absolutely cannot believe in evolution, no matter what. So they automatically must reject all of science, no matter what. And they must automatically claim that they love science and get all of the benefits of science at the same time, no matter what. No free will, no logic, no consistency involved.

At the very least, they don’t seem predestined to change their minds in the face of mere reality.

Of course I think I have free will.

I have no choice but to do so.

Glen Davidson

So in a very real, physical, and non-supernatural sense, your thought processes are not merely the end result of some deterministic brain physics and chemistry.

I wasn’t thinking in terms of physics and chemistry; I was thinking within the concept of a psyche with its foundation in brain tissue.

In the same way that we are not expecting computer chips to make decisions; the decisions are not merely the end result of some deterministic chip physics/electronics.

The decisions are made by the program, in reality they express the programmer’s ideas about what decisions should me made based on the input.

Physical brains or physical computers, they are both tools for processing of data. Although the complexity of the brain goes a bit farther than the computer’s.

Rolf Aalberg said: The decisions are made by the program, in reality they express the programmer’s ideas about what decisions should me made based on the input.

Yes, but I think you’ve gotten to the point where brain-computer analogies fail. In human-designed computers, we try and make the hardware as stable as possible and run the software “on” it. In the brain, the outcome of a software run (which includes environmental inputs etc…) can result in a hardware rewire. In computers, engineers call this a short and work to prevent them from ever happening…and when they do, they try and return the computer to its earlier state. In humans, AFAIK its just part of the regular process of building and repairing neural pathways. Its a feature, not a bug! :)

This is much more likely over time to lead to a form of chaotic development or behavior which won’t be predictable based on the initial state, except in general terms (none of our brains forget how to breathe).

IOW, what you say about computers may be true, but does not apply to us since in important ways, brains aren’t computers.

eric said:

SteveC said:

As a computer guy, I’m sort of baffled by the free will debate. It seems extremely obvious to me that free will, at least as most people think of it, cannot exist

and Rolf Allberg said: Will is not free; it is deterministic. What else could it be?

My latest (and layman’s) understanding of neuroscience is that free will enters the picture because processes do not only propagate “up” (molecular interactions determine neurons determine electrical patterns etc.), they also propagate “down” (your neural pattern can influence your neural structure, which can influence yoru brain chemistry, etc…). So in a very real, physical, and non-supernatural sense, your thought processes are not merely the end result of some deterministic brain physics and chemistry. Rather, they participate in determining how your brain physics and chemistry operates. Your conscious thought is not just an output, its also an input.

This, of course, says little about robots or simulations of robots. :)

Indeed this is an excellent point; and it is a phenomenon that can be observed at nearly every level of complexity.

Emergent properties of complex systems are not generally predictable from “bottom up” analyses of lower level behaviors of system constituents.

The more usual case is that emergent phenomena start to become the dominate influences in further evolution and system behavior. If the underlying “building blocks” (i.e., the fundamental constituents of the system) achieve a rather long-term stability, then these can be treated as fundamental even if they are extremely complex.

It then becomes the interactions among these complex building blocks that set the stage for higher levels of complexity and organization.

At even higher levels of complexity involving complex mixtures of already complex systems behaving in complex ways, the interactions and behaviors involve outputs that become inputs as well as determiners of behavior. This is not limited to brains and complex nervous systems; it happens at much lower levels of complexity in systems that don’t even come close to what we would classify as living organisms.

Kevin B said:

John Kwok said:

I just couldn’t resist:

“Stop, Dave. I’m afraid, Dave. Please … stop.”

I must admit that when I saw “robot” and “will”, I immediately thought of this.…

http://www.youtube.com/watch?v=RG0ochx16Dg

(But was it Dr P Zachary Smith?)

I thought of young Will and the robot too, but I opted instead to quote from Hal.

Not only do humans make plans, restrain instinctive and emotional behaviors, weigh decisions, and so on, but relatively unique brain structures, for example the frontal cortex, are involved in our ability to do so. These brain structures appear to have been strongly selected for in our lineage’s history. Certainly, relative to our primate ancestors and relatives, these anatomic structures are more developed in humans.

Other terrestrial lineages with high “intelligence” and flexible behavior, such as canines and bears, show parallel evolution, but to nowhere the same degree.

We see traits of “intelligence” in less closely related lineages such as some birds, octopi, and marine mammals as well. We do not know whether these types of animals experience a mental state, and currently have no possible means of knowing.

Not only is it the normal human experience to be aware of capacities such as self-restraint and planning, but specific types of brain injuries markedly reduce these capacities. http://en.wikipedia.org/wiki/Dysexe[…]ive_syndrome

I honestly don’t know what “free will” is supposed to mean if it does not mean the accurate perception of being able to modify voluntary behavior on the basis of environmental cues or stored memories.

Anyway, the “free will” issue is probably one of many that is best addressed by using specific language.

great representation of compatibilism. (sarcasm)

best leave the free will debate to those qualified to tackle it. determinism does not mean no freedom or no moral responsibility… the lack of rigor in this response is just awful. I agree with PZ Myers when he defended elitism in academia :)

And the literary allusion I think of when it comes to free will is Douglas Adams Hitchhiker’s Guide.

The full passage is at the link, the short summary: After realizing the that Ultimate Question is possibly stored in Arthur’s brain, Frankie mouse and Benjy mouse try and negotiate the purchase of said brain. Arthur objects, and they offer to replace it with an electronic one. They guarantee no one would ever notice the difference. Arthur objects–he would notice. They explain, no, he’d be programmed not to notice!

stephen said: best leave the free will debate to those qualified to tackle it.

Or maybe best just to leave it alone.

stephen said:

great representation of compatibilism. (sarcasm)

best leave the free will debate to those qualified to tackle it. determinism does not mean no freedom or no moral responsibility… the lack of rigor in this response is just awful. I agree with PZ Myers when he defended elitism in academia :)

Your comment has no apparent meaning and does not address anything that anyone else has said.

… is that the robots apparently made choices and behaved precisely as if they had free will.

1) Which definition of free will are you using?

2) What do you mean by: to behave “as if one has free will”?

It is clear from reading the comments that not everyone is using the same definition, and some participants are very confused as to what the debate is even about. It is pointless to get involved in a back-and-forth argument until definitions have been established. Until there are definitions, the debate isn’t “about” anything meaningful.

eric said: My latest (and layman’s) understanding of neuroscience is that free will enters the picture because processes do not only propagate “up” (molecular interactions determine neurons determine electrical patterns etc.), they also propagate “down” (your neural pattern can influence your neural structure, which can influence your brain chemistry, etc…). So in a very real, physical, and non-supernatural sense, your thought processes are not merely the end result of some deterministic brain physics and chemistry. Rather, they participate in determining how your brain physics and chemistry operates. Your conscious thought is not just an output, its also an input.

Now here’s an example. What does eric mean by free will? Does he mean lack of determinism or does he mean lack of predictability?

If the former, I think he is wrong. At what point during that up-propagation and down-propagation is causal determinism violated? The laws of physics and chemistry are being followed at every step. The systems which do the up- and down-propagation are certainly, in the views of a naturalist/materialist, just part of the material world.

A test for free will: choose not to have your next thought.

wamba said: What does eric mean by free will? Does he mean lack of determinism or does he mean lack of predictability?

Neither is exactly right. I mean the pattern(s) that make up your consciousness actively participates in determining what you do next.

I think we get too easily stuck in simplistic philosophical models of things where arrows go one way; A leads to B leads to C, and boostrapping is nonsensical; A must have a non-A cause. But that is not how the real world works. Sometimes A leads to B and B leads to A, and they do iteratively bootstrap. Sometimes, its even more complicated than that. In these sorts of cases, asking whether A is a “cause” or “effect” becomes meaningless. It becomes a chicken and egg joke routine.

If electrical patterns of activity (consciousness and subconsciousness) influence brain strucutre and vice versa, we have a chicken-and-egg routine. There is no longer any real point in asking which is the true cause and which is merely the deterministic effect; both are both.

The laws of physics and chemistry are being followed at every step. The systems which do the up- and down-propagation are certainly, in the views of a naturalist/materialist, just part of the material world.

I agree wholeheartedly.

Does he mean lack of determinism or does he mean lack of predictability?

People often confuse determinism and predictability; they are not the same. When I said free will, I meant lack of determinism. That is, it is hard to see where free will comes from, unless humans are somehow immune to the laws of physics and chemistry.

wamba said:

Now here’s an example. What does eric mean by free will? Does he mean lack of determinism or does he mean lack of predictability?

If the former, I think he is wrong. At what point during that up-propagation and down-propagation is causal determinism violated? The laws of physics and chemistry are being followed at every step. The systems which do the up- and down-propagation are certainly, in the views of a naturalist/materialist, just part of the material world.

If by “determinism” you are referring to the laws of chemistry and physics, there is no evidence whatsoever that the laws of chemistry and physics are being violated in any way.

In fact, things like the autonomic nervous system, and the fact that taking the nervous system out of a very narrow temperature range shuts it down at low temperatures and sends it into chaos at high temperatures, is a pretty good indication that physics and chemistry are working just fine.

So it has to be connected to the complexity and organization within the system. An approach to this problem would start with extremely simple nervous and sensory systems that simply orient an organism toward food or energy. Such systems are fairly easy to understand; and humans have been building such things in the way of mechanical and electromechanical systems for a very long time now.

Now add rudimentary memory that stores previous “successes” at finding “food” or energy. In such cases, the sensory and nervous systems are now augmented with a means of operating passed long delays between the acquisitions of enough sensory stimuli for the nervous system to use.

Expand the memory to include “meta-memory,” i.e., memories of memories. Somewhere in this process the sense of time evolves within the system because such memories can now have memories of events that are NOT followed by other events along with memories of those same events followed by “later” events, and finally by memories that include whole sequences of events and how such sequences relate to other sequences and “when” they appeared in memory relative to those other sequences.

Much of the problem in analyzing the behaviors of such complexity is that there is so much “parallelism” in the activities going on within. Trying to make comparisons between systems like this and the sequentially operating configurations of today’s computers is not going to be very helpful. We really need more experience with complex parallel computers that actually function with suitable “inhibitors” that prevent clashes of instructions at any given node.

Nervous systems appear to be highly and delicately coordinated systems operating as a unified whole. That’s part of what is currently under study; we don’t yet understand how such systems behave.

But somewhere out of that coordinated “mess” must emerge the notion of consciousness and the feeling of willing things to happen within the limitations of the system.

Nice info :-) visit my blog pfllc.blogspot.com

About this Entry

This page contains a single entry by Matt Young published on May 10, 2011 9:25 PM.

Photography contest, third edition? was the previous entry in this blog.

Exploring the protein universe: a response to Doug Axe is the next entry in this blog.

Find recent content on the main index or look in the archives to find all content.

Categories

Archives

Author Archives

Powered by Movable Type 4.381

Site Meter