Do humans have free will?
32
789
2053
25%
chance

One of the biggest assumptions our society is founded in is that humans have, more or less, direct control over their actions. But science has in recent decades been winnowing the potential mechanisms for this to operate, suggesting more and more that humans are fleshy algorithms who respond to stimuli in complex and interesting but ultimately automatic ways.

Market runs for a long time because it seems like this might not be resolved for a good long while. I expect this market to extend if resolution is not clear at the end date.

Resolution criteria for this is obviously hard as it's going to depend on as-yet unknown evidence, so I'm going to float a few suggested resolution criteria and open the floor to more suggestions.

Resolves NO if an AI is able to reliably predict human responses to stimuli 99.9% of the time ten or more minutes into the future in a variety of situations.

Resolves YES if a mechanism explaining free will is found within human biology that defies prediction through biological process.

Resolves as per unanimous agreement of a council of interdisciplinary scientists (e.g. physicists + neurobiologists + philosophers) of international repute.

Disclaimer: whether or not free will is real (and even a 0.01% chance of YES means you should not assume NO given the impact either way), feeling like you have free agency is important for your mental well-being. If you feel you have no free will then I'd recommend seeking help from friends or therapists. Also, this market favouring NO does not absolve you of responsibility or the consequences of your actions, you are not merely a ball on the pinball table of an uncaring, pinball-playing good and someone somewhere will care if you think that you are.

Get Ṁ600 play money
Sort by:
predicts NO

OK, I’ve read / watched the points made below and have come to the following conclusions:

1) I will not be re-titling the question. It’s a question people will want to see asked and answered, and if I add a bunch of qualifiers to the question then it will A) put some people off getting involved and B) lead someone else to ask it in turn. In fact it turns out there is already a “do we have libertarian free will” question which search just plain did not show me and has barely any responses, which reinforces my decision to leave the question big and broad.

2) It’s a very reasonable point about the need for the belief in free will to ensure mental well-being and I will add a disclaimer as such to the question body.

3) there are valid points about the criteria being overly literal, but consider the counter-factual: if humans are entirely deterministic and free will is “just” the story our brain tells us while doing things, I don’t think many would argue that was “true” free will. A discovery as such would also have implications that would need guarding against: aside from impact on mental well-being, it would imply it may be possible to compel people to do things which they believe they are choosing to do and are part of the core of their being. For an app full of AI doom and alien invasion markets I don’t think it’s unreasonable for this to be weighed in here too.

I’m open to having my mind changed on point 3 and the main question body, and given this question runs for many years I’m not going to be hanging on for new comments, so if you feel I’m plain wrong about anything then take your time and write it out clearly and with actionable changes to the question summary, and I’ll consider them as thoughtfully as I can. And if you don’t like the formulation at all and think I’m unreasonable then you are of course free to not bet and to write your own question.

@Noit The trouble is that you aren't keeping the question "big and broad". Based on the current description, this question is only about libertarian free will, so titling the question as if it's about free will more broadly is misleading. If you're just asking, "Do we have free will?", you have to include compatibilism as a possibility. Compatibilism is actually the position held by the majority of philosophers, but this market pretends it doesn't even exist. So yes, many do argue that this is "true" free will.

Also, I would not say free will is "just" the story our brain is telling us. It is an actual state of affairs that really does hold, not just some cognitive illusion. Your conscious mind (i.e. you) really does cause your actions and decisions through deliberation, and you're not forced to make a decision by some external factors. That is what most people mean by free will, and it doesn't conflict with determinism in any way. So what if the algorithms you use to make decisions are determined by the laws of physics? That doesn't mean you don't have control over them. Our minds are physical processes that exert control through physical means. If our decisions weren't determined by the laws of physics, that would mean we don't have control over them, since it would imply that our actions are either random or controlled by an external non-physical factor. So under my view, determinism to some extent is actually a necessary precondition for free will (I say "to some extent" because it's not required for determinism to hold universally. As long as our actions are mostly determined by physical laws, so that our conscious algorithms can control them, that is sufficient for free will even if there is some small quantum-scale indeterminacy).

predicts NO

@JosephNoonan I mean, one of the resolution criteria is interdisciplinary agreement of a panel that includes philosophers, so if they all came out and said “yes free will compatibilism is overwhelmingly and conclusively our state of being, we have free will” then this would resolve YES.

But as I’ve literally been introduced to compatibilism by your response, I obviously have some reading to do. Call me uneducated if you will but I suspect I’m in the majority in ignorance of it, and I think it’s healthy to have the big question with stuff like this in the comments to drive people towards more specific truths.

Again, feel free to make a question you think is better and link it here.

predicts NO

@Noit Or to put it another way: what resolution criteria would you like to see added, or changed?

@Noit Well, if you are going to include compatibilism, you would have to change a few things in the description:

science has in recent decades been winnowing the potential mechanisms for this to operate, suggesting more and more that humans are fleshy algorithms who respond to stimuli in complex and interesting but ultimately automatic ways.

This only applies to libertarian free will, for example. Under compatibilism, "fleshy algorithms" is exactly what we would expect humans to be. In fact, I would argue that science has already discovered many of the mechanisms by which free will operates. We call them "psychology".

I know that the resolution criteria you had here were just a suggestion, but they don't work even in principle if you're trying to include compatibilist free will.

Resolves NO if an AI is able to reliably predict human responses to stimuli 99.9% of the time ten or more minutes into the future in a variety of situations.

Resolves YES if a mechanism explaining free will is found within human biology that defies prediction through biological process.

As a compatibilist, I expect that an AI could, in principle at least, perfectly predict human actions if given enough information about the laws of physics, the initial state of the person's brain, and the situation the person was in. In fact, on my view, the AI itself could have free will. And I would not expect a mechanism within human biology that defies prediction to be found because I don't think that's required for free will. In fact, if such a mechanism was an integral part of our decision-making process, that would actually be a problem for me, since it could imply that our decisions are random rather than determined.

As for what resolution criteria you should use, the only viable option I can come up with is "Resolves when there's a consensus of philosophers." It can't resolve based on a scientific experiment because compatibilism vs. incompatibilism isn't a scientific question. Under a compatibilist view, free will has already been proven to exist, since all you need to do to show that it exists is find a conscious decision-making process operating correctly without interference, and then show that it actually causes the decided-upon action to be carried out. On the other hand, libertarian free will is all but scientifically ruled out already, and will probably be completely ruled out by science eventually (assuming humanity last long enough and develops neurology well enough to do that).

And yes, I know that's not a very satisfying criterion, since it's almost impossible for philosophers to reach a consensus on something like this. That's why I originally suggested changing the name of the question. I don't think doing so would limit the reach - the other question you mention is presumably unlisted, which is why it has so few traders - but if you want to keep this question broad, you will have to settle for a more nebulous/harder to satisfy criterion, since you will have to rule out more alternative options before resolution.

Call me uneducated if you will but I suspect I’m in the majority in ignorance of it

Yes, this is why I said below that the popular framing of the debate is so terrible. Most people don't even know that the majority position among philosophers exists!

predicts NO

@Noit, @JosephNoonan

if humans are entirely deterministic and free will is “just” the story our brain tells us while doing things, I don’t think many would argue that was “true” free will.

This is more or less exactly what I am arguing: a "true" free will exists (often) for me that is 100% compatible with a fully deterministic reality. This is because it comes from "me", a very important agential construct that exists within the doors of my perception: "I" lives within my dream, not within the particles themselves that generate everyone's dreams. Looking for the subjective self in the particles/biology only .. gives us very odd conclusions -- it's the wrong dimension.

"I" is absolutely not 100% solid, for example I meditate a TON so I'm constantly aware of how my sense of "being a singular node of agency" can become stronger and weaker at different moments, disappearing all-together often, with a pervading oneness being the "outer context" that I easily dissolve into when I allow these DMN brain circuits to rest. So I'm very sympathetic to the argument that it's not "true" in the sense of being some hard-core solid thing that exists always and you can always push against. But crucially, overall, I have a great level of direct influence on the manifestation of my life, along with a need to frequently make personal choices that "I" must own and live with the consequences of.

it would imply it may be possible to compel people to do things which they believe they are choosing to do and are part of the core of their being

I see this as not just possible, but something that happens often in our world -- political manipulation, for example. Make them think it's their own idea, even identity -- classic technique.

Personally, I think there are a ton of very interesting questions to ask in this space, including the market description's current criteria. We don't know what base reality even is, for example, or the dream, and their relationship. But because the term "free will" has meanings with such different and conflicting semantics, it's crucial that we use specific language when probing the space. Unfortunately I think most of the conversations people have about these concepts is confused heavily by the conflation of these different meanings and a lack of language for surrounding concepts, and I'm concerned this market in its current state may add to that.

I think a lot of the confusion people have on these topics just comes down to a lack of clear insight and language to pick apart the different aspects of the paradox of being both an individual and one-with-the-world at the same time, and the properties of various aspects of that... rather than something that humanity hasn't already figured out.

As far as the title question of "Do humans have free will" -- I would say that simply depends on how one chooses to define the phrase, rather than on any questions about the nature of reality.

Discussions like this miss the point entirely IMO -- it's an overly-literal interpretation of what the phrase really means.

You don't find free will in particles, or biology -- you find it in the dream of being an individual. It's simply the self-recognized agency re: modeling the nature of there being a locus of control associated with your personhood in the world.

Josha Back describes this well here (from 46:02 to 50:08 in the podcast for a short elucidating example, also see the deeper dive from 19:51 to 33:56). Here's add'l thoughts of his on the meaning of free will from a more psychological perspective.


TBH, I think trying to define free will in an ultra-literal manner is quite silly and can actually cause harm.

Silly because of course we don't find free will in particles. It may be "only" an abstraction that minds make about the world/themselves to navigate appropriately, but try living without that abstraction. It's a completely essential abstraction to have, fundamental to being a coherent agent of any type.

Harm because ultra-literal definitions are so different to what most people actually mean when they use the phrase, it can cause confusion and even a sense of helplessness when people hear "such and such expert denies the existence of free will" -- when what's really going on is the "expert" is talking about something else entirely from the common definition.

For example, from Claude's summary of this study:

  • Participants' definitions of free will emphasized the ability to make choices and control one's actions, consistent with previous research.

  • The associations between FWB [Free Will Belief] and subjective well-being appear to be due to covariation between FWB and sense of control/choice.

  • Belief in the ability to control one's life and make choices may be protective against stress and depression.

  • The predictive utility of FWB for personal life outcomes like well-being seems to be explained by perceptions of personal control and choice.

This is how most people use the phrase, and it's very important that we perceive and emphasize this notion of having choice and control in our lives in order to be successful at living high-quality lives.

I don't think this market calls into question as much substance as it might seem to. I would suggest renaming it to something along the lines of "Can free will come from raw physics/biology alone?"

@TomPotter I totally agree.

Participants' definitions of free will emphasized the ability to make choices and control one's actions, consistent with previous research.

And this is why the way the free will debate is framed for popular audiences is so terrible. It's always framed as free will vs determinism, with free will being some seemingly magical dualistic force that would allow you to violate the laws of physics. This in no way comports with what anyone actually means by free will, and this framing of it has mislead so many people either into thinking that science proves we don't have free will, or that free will does exist and really is this magical physics-defying ability.

this framing of it has mislead so many people either into thinking that science proves we don't have free will, or that free will does exist and really is this magical physics-defying ability.


@JosephNoonan Exactly. It's a real problem.

predicts NO

I’ve just woken up and haven’t time to properly consider this now, so I’m not ignoring this but need some time to follow your links and review my thinking.

predicts NO

[Updated topic post to include an add'l clip of Josha's w/ a more psychological perspective]
[EDIT: oops broke the link doing so, fixed now.]

You should change the title to "libertarian free will". I have a very high credence that humans have free will, but I would never bet YES on this market because the resolution criteria have nothing to do with whether humans actually have free will - they are actually about whether human actions are determined.

predicts YES

@JosephNoonan They are not; assuming that compatibilism is true, it is still the case that no AI, or anything else, will be able to predict human responses with 99.9% accuracy.

For one thing, humans are capable of following the algorithm "do the opposite of what you predicted," and that is enough to avoid being predicted with any degree of accuracy.

Note that even 100% deterministic machines are also capable of following that algorithm, so the resolution criteria have nothing to do with determinism.

@DavidBolin The market doesn't say anything about the AI having to tell the human in advance what they have predicted. Yes, it's impossible for an AI to predict a human's actions if the AI is planning to tell the human what it predicted and the human is planning to do the opposite, but it's perfectly possible for an AI to predict what a human will do if it isn't going to interact with the human, or at the very least won't change how it interacts on the basis of its prediction. It's the interaction between the AI and the human that spoils the prediction, since the way that the AI will interact with the human depends on what it predicts, but what it should predict depends on how it interacts with the human. The market currently does not say that there needs to be any such interaction, so the criteria don't work.

how would it resolve if this is proven to be unknowable?

predicts NO

@Tater If it was a Heisenberg cannot-know-x-and-y way, i think it’d be a YES resolution because if we assume fundamental unprovability then society as a whole is not going to be redesigned around the negative. We’ll continue to act as if we do and so a duck test would say that we do. But I wouldn’t resolve immediately on a single paper saying as such about a single approach, I’d want consensus on “this is unknowable” to form.

If it was a “we’re all out of ideas” or an “an experiment could be designed but requires infinite resources” then I’d probably extend awaiting other approaches and ideas.

@Noit Obviously people's actions are not openly predictable, since people are capable of using the algorithm "if they predict I am going to do A, do not A."

Just to check this is AI predicting the acts of an individual, as opposed to groups/society. In groups this isbthe same as Hari Seldons Psychohistory as per Asimov's Foundation as an evolution of sociology. Which i think would be much more likely to become true (predictable by AI/quantum compuer), than the action of an individual.

predicts NO

@DanielPugh Yes, this is purely about individual free will. The experiment should be designed in such a way as to evidence that an AI can either identify e.g. the answer to a question before it is asked, or even prime a subject to behave in a certain way in response to a stimulus.

I agree that the Foundation-style “we can broadly predict what society will do next” is probably a simpler problem because it would never have to deal with minor “do I walk or get the bus “ type questions.

Cool, understood!

“if an AI is able to reliably predict human responses to stimuli 99.9% of the time ten or more minutes into the future,” we have basically achieved future time travel.

The outcome is deterministic and therefore has nothing to do with probability. My bet was inevitable.