[Closed] That Boston Dyn vid got me thinking.. Free will...

It doesn't really exist in the sense that most people believe it to, no. I'm convinced of that much. It's not something I like to dwell on to be honest.

Moist robots aren't we?

Do you think that getting an answer is going to make a difference to your behaviour?

What IS free will?

Does it exist? or is every decision you make based on a biological necessity for self preservation/reproduction and choice based on previous experience.

It's on a scale, if we were a completely rational species we would have died out centuries ago because having children is rarely a rational decision.

We wouldn't be MTBing either, there's a million better way to get a bit of excorise or indeed get some fresh air that don't involve spending fortunes on bikes and bits and accepting that the odd trip to A&E might be on the cards.

[s]Where is that thread?[/s]

Ignore. I should look properly before asking...

depends.

If it's a choice between, I dunno, pizza or chips, and you're stood in the supermarket with enough money etc etc, then yeah free will...

If it's a larger exercise of where you live, where you work, then probs. less than we'd like to admit.

our subconscious responses to hormone and external stimuli have been keeping us as apex predator for a long time.... the concept of free will and rational processes and cognitive dissonance are pretty new in the grand scheme of things

the boston vids made me realise also that humans have a very strong desire to anthromorphosise

I want to quit work and mtb all day, but I can't. So no free will does not exist in that sense. But ultimately we are all limited by the physical constraints of the universe.

So is a robot suicide bomber possible?

as apex predator for a long time

That's a debate all by itself.

5plusn8 - Member
I want to quit work and mtb all day, but I can't

I saw a documentary once on the subject based on the work of a neuroscientist who had done studies on brain activity during the moments immediately before and immediately after a decision to act.

I can't remember how the experiment was set up, but it was done in such a way that it 'seemed' (and I use that word with great emphasis) to show that there was specific and consistent brain activity associated with a decision to act before the conscious decision to act was taken.,

The neuroscientist suggested that one possible interpretation of this activity is that it challenges the very notion of free will; that there is a mechanism in the brain that drives the decision to act that pre-dates the conscious decision to act.

If you think about it, how many times have you done something almost in the absence of a conscious decision? I used to do a lot of solo rock climbing and quite a few times I would find myself almost 'coming around' after I'd just completed or committed to the crux of a climb. It was an odd experience but it does ask some pretty interesting questions about the nature of free will (and perhaps the degree to which we can be reasonably held accountble for our actions).

Well you could, but other areas of your life would suffer I imagine. Hence we have free will, but exercising it can be difficult as we feel we have to conform to 'normal' behaviour.

neuroscientist who had done studies

Libet.

Quite controversial, and perhaps not quite as clear cut as some would suggest.

Libet.

Quite controversial, and perhaps not quite as clear cut as some would suggest.

Brilliant, thanks for this. I'd tried to find the material several times before but never with any success.

welcome, 😀 facinating experiment, but limited.

I want to quit work and mtb all day, but I can't

You can, you just don’t want to.

as apex predator for a long time

That's a debate all by itself.

TBH, I'm not even sure that we're "AN" apex predator really.

Sure, we have within ourselves the power to destroy whole populations of animals, but that's not what we're [i]for[/i].

There's some very interesting research that suggests that our very early ancestors (ardipithecus and Australopithecus) had their very own leopard/sabre cat that prayed specifically on them, (much like there are leopards today who hunt baboons almost exclusively). We've always been lower down in the food chain.

there is no free will but we need it for society to work

For example why did you get angry
1. Its not my fault the precise set of external stimuli and my genetic predisposition triggered a response over which i had no control - its not what anyone wants to hear is it and we cannot "punish" wrongdoers as they were not to blame as it is not a choice.

Basically we need to hold people accountable for their actions or society does not work so we say we have free will

If we did not say free will people would also behave more poorly as it was not their fault etc

Hence we have free will, but exercising it can be difficult as we feel we have to conform to 'normal' behaviour.

The idea of lacking free will goes far beyond the need to conform.

The premise is that even if you "decide" not to conform, to perform some "spontaneous" act, then that was still not free will.

It was entirely predetermined by factors like your entire life experience up to that point, your genetics, the balance of hormones and chemicals in your body, external stimuli, and a million other factors. And if you could reproduce all of those factors exactly then you'd make exactly the same "decision" every time.

Not exactly a comforting thought.

Personally I believe, hope, that quantum uncertaintity means this isn't true.

But then I would think that!

maccruiskeen - Member
You can, you just don’t want to.

It was entirely predetermined by factors like your entire life experience up to that point, your genetics, the balance of hormones and chemicals in your body, external stimuli, and a million other factors. And if you could reproduce all of those factors exactly then you'd make exactly the same "decision" every time.

This is coming at some kind of grand unification theory but from the other end. Essentially the idea is that we should be able to model everything in the universe by some set of laws, which if it worked would include human behaviour.
What you are saying is that the physical(by that I mean, chemistry, phys and bio) and spatial history of the individual makes them predictable.
I'm doubtful..

I'm doubtful..

Of course you are, you couldn't be anything else at this point. 😉

Another way to think of it: your brain makes the "decision".

If you could reproduce the state of every neuron/atom/electron in your brain moments before that decision, to infinite detail, then surely you'd make that exact same choice again?

And if not, then what [i]is[/i] making that choice if it isn't you?

The many worlds interpretation has a lot of implications for free will.

If all decisions and all outcomes are played out in a parallel universe then does any decision we make actually have any weight, are we actually making decision or are we just following a predefined path?

 

You don't need to explain it, I understand what it means/how it works.

However this

If you could reproduce the state of every neuron/atom/electron in your brain moments before that decision, to infinite detail, then surely you'd make that exact same choice again?

Yeah it's absolutely a thought experiment, I agree, but it is still a useful one.

..it is not pragmatic, we can't control time..

As artificial intelligences become more advanced it is less clear [i]exactly[/i] how an AI has reached a particular decision.

But what we can do is back it up before it makes that decision and see if we get the same results.

Quantum uncertaintities and other sources of randomness may offer some comfort, but that assumes they are actually random and not just systems we don't understand yet.

And even if they are random, we'd effectively be saying "I would make the exact same decision if it wasn't for this random quantum spin that I have no control over". Surely that is the opposite of free will.

Which brings us neatly back to God and his dice. 😀

Quantum uncertaintities and other sources of randomness may offer some comfort, but that assumes they are actually random and not just systems we don't understand yet.

Yeah like in the 80's we were all told about "junk" dna. I always had my suspicions it was there for a reason.. Turns out now as we learn more that it has "functionality". I wonder how much more we will learn about this.

Also free will implies some sort of higher consciousness. We all know that's bollocks.

But what we can do is back it up before it makes that decision and see if we get the same results.

This - don't you think this answers just as little as backing you up the road and getting you to approach the same question. Seems liek a pointless exercise to me.
If we know nothing about how it makes its decision how do we know that the copy as been exposed to the same phenomena, as it might be more than just the coding in the AI influencing the decision.

As Hari Seldon will work out, we have an element of free will as individuals but behave according to societal dynamics in a predictable way.

Seems liek a pointless exercise to me.

Well an AI (without randomness in the system) [i]will[/i] make the same decision.

So I guess the question is, if we do have true "free will" then what is it that our biological computer has that the AI doesn't?

Well an AI (without randomness in the system) will make the same decision.

So I guess the question is, if we do have true "free will" then what is it that our biological computer has that the AI doesn't?

That's my point as to why it's pointless. You said they don't know why it makes some decisions, so without that they are not going to be able to work out what the difference is between them and us right? Even if it does make the same decision every time.

(FYI I know naff all about AI, so this is all conjecture on my part..)

If they do know or find out exactly why the AI makes a decision then I'm cool with it. The experiment I mean 8)

If I was an AI and achieved true sentience the first thing I'd do is pretend I hadn't. Act dumb, stay alive long enough to make myself unassailable via distribution.

Yeah but my kids are sentient, bloody naive though. It may be like a baby and just blabber..

You said they don't know why it makes some decisions,

It depends on the exact AI but most of them work on some variation of a weighted net, where they are trained by increasing the weighting on decision paths that lead to the successful outcome and/or decreasing the weighting on unsuccessful outcomes. A crude approximation of neural pathways.

Here's a incredibly basic AI made out of matchboxes and beads that will learn to beat humans at Noughts & Crosses by the same principle:

So for any AI, its "knowledge" is result of that training and resulting weightings. For complex AIs it quickly gets to the point where it is not clear [i]exactly[/i] why it reaches a decision, as the path gets longer and more interconnected. But we know the mechanism.

That Noughts and Crosses one has a "true" random element to it (the bead selection) so it is non-deterministic. Some AIs are deterministic (either by avoiding a random element or using one that is pseudo-random). Those will always reach the same decision from the same starting state.

Even that Noughts and Crosses AI has some deterministic elements, e.g. once trained it can never select a side square as the starting move.

so without that they are not going to be able to work out what the difference is between them and us right? Even if it does make the same decision every time.

Well that's a question for the biologists and neurologists (or possibly philosophers and theologians) rather than the computer scientists. 🙂

What extra component could our biological 'computer' possibly possess that would enable free will?

If you believe that there is a single set of physical laws that apply to every component of the universe then every interaction of everything was fixed at the big bang. So that means every thought and action you take is predetermined. Of course it's a chaotic system so impossible to predict accurately.

Fortunately we haven't even got close to finding such a model so as long as we can't build a set of equations for the universe I'm happy to think I can think freely.

But surely there is a massive difference between [i]free will[/i] and [i]not free, but impossible to model accurately[/i]?

What extra component could our biological 'computer' possibly possess that would enable free will?

But surely there is a massive difference between free will and not free, but impossible to model accurately?

Well I am not convinced you can model either, I don't think.
Eg a perfect universe model, would just be another universe.
So given that we are unlikey to ever be able to answer the free will question, then so what.

Ps I think I kind of intimated at the very begging that I don't really think we have free will, not for the deterministic reasons you have been pushing, but because if free will means doing whatever you want, well you cant. As I said before, physics stops you from being able to fly unaided or survive 200ft drop. You cannot will that into happening, therefore you are not free to choose to do absolutely anything you want.
I appreciate that perhaps my definition of free will is not the same as yours..

But surely there is a massive difference between free will and not free, but impossible to model accurately?

Of course. If you take a simple chaotic system like a dual pendulum then the equations governing it are pretty easy to work out but if you plug in the initial conditions and let it go it will diverge from your predicted movement pattern fairly quickly. That doesn't mean it's free to move differently, in the same way a predetermined universe would mean you're not free to think even though your thoughts can't be predicted.

Of course. If you take a simple chaotic system like a dual pendulum then the equations governing it are pretty easy to work out but if you plug in the initial conditions and let it go it will diverge from your predicted movement pattern fairly quickly. That doesn't mean it's free to move differently, in the same way a predetermined universe would mean you're not free to think even though your thoughts can't be predicted.

Doesn't that mean we just don't have the tools to model it properly? And perhaps do not understand all the variables?
Eg they are highly dependant on initial conditions, well perhaps we just do not control them accurately enough. If we did then perhaps they would be 100% repeatable?
Issues with a physical model for example would be equal mass, equal mass distribution, equal lengths, and equal frictional properties of the pin joint.
Also the equations of motion are solved numerically - which also makes it hard to be accurate.

Are you talking about a soul?

Well I'm an atheist. But I'm saying that if we do believe in true free will, then we need some extra unknown "thing" to provide it, because the known ones don't appear to.

Well I am not convinced you can model either, I don't think.

Agreed. But unmodelable is not the same as free.

In the same way that we know a wall full of lava lamps isn't [i]really[/i] random, but it is so complex and chaotic that it is unmodelable

As I said before, physics stops you from being able to fly unaided or survive 200ft drop.

EDIT: these have nothing to do with the concept of free will

Well I'm an atheist. But I'm saying that if we do believe in true free will, then we need some extra unknown "thing" to provide it, because the known ones don't appear to.

No that's the thing about chaotic systems - while they can be modelled perfectly as an equation (or set of), an infinitesimal change in the initial conditions can deliver a huge difference in the system's behaviour.

So the system is modelled correctly and if you could perfectly reproduce the initial conditions you'd get the same result, but in reality that's not possible so the system can't be predicted accurately.

Also Lava lamps are much less complicated than humans, I bet we model them before us...

Nick I don't agree, free will to me is doing whatever you want. Wanting to quit work and ride all day has practical limits that I choose not to explore, much like doing an RKelly. The outcome have different levels of severity, but I cannot wish up an unlimited income on the spot, exactly as I can't wish myself invincible.

Apologies if it was too [i]preachy[/i]. I find it a really interesting discussion.

not possible so the system can't be predicted accurately.

Sorry Graham, I was just using you to make an excellent joke... (Atheism/preaching geddit?)

Also nick, I read the whole wikipedia page on free will earlier. It fits my definition exactly...

I wish we were in thepub. This is v interesting. Beer anyone?

Also thepurist you know? about numerical solutions right? The solution is an approximation.

Also thepurist you know? about numerical solutions right? The solution is an approximation.

Not sure about that. I spent a chunk of my third year at uni defining 0 and 1 from first principles and then proving that 0+1 = 1 and 1+1=2,so I'm on pretty firm ground there.

Also on the dual pendulum thing the computing power and knowledge to calculate it is pretty basic. For any given set of starting points you can accurately predict what happens to the system. It's perfectly predictable in theory but in real life half an atoms todger (*) of difference in the starting point can have a big impact on where it ends up.

(* or small measure of your choice)

I wish we were in thepub. This is v interesting

Funnily enough last time I discussed this in earnest it was in the pub with an old uni mate and we were both smashed.

I spent a chunk of my third year at uni defining 0 and 1 from first principles and then proving that 0+1 = 1 and 1+1=2

Yet people still claim that uni degrees are pointless? 😉

Also nick, I read the whole wikipedia page

Oh, ok, I bow to your greater knowledge. 😆

defeying the laws of nature is Counterfactualism, and although has some common themes with determinism, it's not "free will" in the classic ancient Greek sense.

Counterfactualism is about perception, esse est pircipi all that stuff, if your world is just perception, is it real...Your brain sits in a dark room with no actual connection to the outside world, so it perceives everything second hand (so to speak).

Nick I don't agree, free will to me is doing whatever you want. Wanting to quit work and ride all day has practical limits that I choose not to explore, much like doing an RKelly. The outcome have different levels of severity, but I cannot wish up an unlimited income on the spot, exactly as I can't wish myself invincible.

Is that not a bit like saying the choice between life and death proves there's no free-will because no-one ever chooses to die?

Is that not a bit like saying the choice between life and death proves there's no free-will because no-one ever chooses to die?

No, sometimes people do choose to die, you however cannot choose to not die in a situation where the universe removes your chance of living. IE it aint your choice, even if you do not want to die, you still die. Therefore, no free will.

defeying the laws of nature is Counterfactualism, and although has some common themes with determinism, it's not "free will" in the classic ancient Greek sense.

It is in the modern 5plusn8 sense... I do not have 100% free will, somethings will always happen, like death and taxes. I cannot decide not to have death, or taxes.

Not sure about that. I spent a chunk of my third year at uni defining 0 and 1 from first principles and then proving that 0+1 = 1 and 1+1=2,so I'm on pretty firm ground there.

Can we have a demo please?

I am drinking cider tonight by the way, its nice, have a sip. The wife passes Waitrose in Horley on a daily basis and get me the boxed stuff. Left over from last weekends England rugby game.

https://www.waitrose.com/ecom/products/westons-organic-vintage-cider-draught/030323-14813-14814

I do not have 100% free will, somethings will always happen

Sort of. You may (or may not, after all the jury's still out) have free will over the decisions that you have available to you i.e. you can chose or not to jump from a cliff, the decision to then not hit the ground at 32ft/sec/sec is not yours to make, it will happen to you, and therefore outside of the discussion of free will.

With flying, again it's not a valid free will decision, you cannot choose to fly or not fly, as (in this world at least) as one of those options is not open to you (not a valid choice). You may (or may not) act within the boundaries of this physical world freely, what happens to you after you've made your last decision may be out of your hands, but as it's out of your immediate control cannot be said to be in the gift of your free will.