Last year I published an article in The Philosophical Forum entitled, “A New Theory of Free Will” (a free unpublished version is available here). Because I think the theory the article develops may be of interest to readers of this blog who may not have come across it, I thought I might provide a brief précis of the article here, as well as lay out some additional evidence for why I think the theory is worth taking seriously.
Nick Bostrom is of course well-known for arguing, on probabilistic grounds, that we are probably living in a simulation. Somewhat similarly, David Chalmers has argued that we should consider the “simulation hypothesis” not as a skeptical hypothesis that threatens our having knowledge of the external world, but rather as a metaphysical hypothesis regarding what our world is made of. Finally, the simulation hypothesis is gaining some traction in physics.
My 2013 article and subsequent unpublished work go several steps further, arguing that a new form of the simulation hypothesis -- what I call the Peer-to-Peer (P2P) Simulation Hypothesis -- is not only implied by several serious hypotheses in philosophy and physics, but that it also provides a unified explanation of (A) the mind-body problem, (B) the problem of free will, and (C) several fundamental features of quantum mechanics, while (D) providing a new solution to the problem of free will that I call "Libertarian Compatibilism."
I begin my 2013 paper by arguing that we presently have some philosophical and scientific evidence in favor of each of the following hypotheses:
- Eternalism about physical objects and properties: past, present, and future physical objects and properties exist “timelessly.”
- Mind-body dualism: the mind is at least partly comprised by non-physical properties or substance(s).
- Subjectivity about the passage of time: time’s passage occurs within consciousness alone.
- Only one timeline (ours) is actualized, or consciously experienced by observers.
- The holographic principle: the physical universe is simply a series of ordered two-dimensional information (i.e. 1's and 0's) “written” on the cosmological horizon.
- The multiverse hypothesis: the observable physical universe is merely a small part of vast multiverse of alternative possibilities.
I then argue that if all of these hypotheses are true, it follows that our reality is structurally identical to a particular kind of online interactive computer simulation: a peer-to-peer networked simulation.
In computer science, there are two types of online simulations: (A) dedicated server simulations, in which one computer on the network (the "server") serves as the definitive representation of objects and properties in the simulation, and (B) peer-to-peer networked simulations, in which no single computer on the network serves as the definitive representation of objects and properties in the simulation, but in which, instead, "the" simulation is simply represented in parallel on the various interacting computers on the network.
Here is why, if all of the aforementioned hypotheses are true, our reality is structurally identical to a P2P simulation. A P2P simulation just is:
- An array two-dimensional information (e.g. each computer's game program or DVD)
- Comprising a vast array of "possible pasts, presents, and futures" for the simulation,
- Being read in real time,
- By a multitude of external measuring devices (i.e. each computer on the network),
- All interacting in parallel, such that
- The joint measurements of all the computers on the network result in the appearance of single observed, intersubjective reality.
These six features of a P2P simulation are identical to the six aformentioned hypotheses.
Next, let us think about the philosophy of mind and the problem of free will.
First, the P2P model explains, and promises a new resolution to, the problem of free will. Here's how. Anyone who has played an online simulation knows that once one “plays the game”, one can rewind the game back to the beginning, press the “play” button, and watch all of the game that just completed inexorably play out just as it did the first time. Accordingly, from the reference-frame of anyone within such a simulated reality, their “physical” world would appear to be causally closed. Observers within the system, when looking at the physics of their system (viz. at how gravity works in the simulated environment, etc,) would be led to believe that the system must be closed under the laws of their physics. But, of course, we – the users, outside of the simulation – know that their experience of causal closure in their reference frame is an illusion generated by the choices we freely make in our higher reference-frame (where our choices "outside" the simulation dictate what "inexorably" happens within it). Finally, although in our case our choices outside of online simulations appear themselves to be dictated by the laws of physics, the model makes room, in principle, for true libertarian free will in a higher reference-frame -- viz. pure conscious will (if mind-body dualism is true) -- to generate the illusion of causal in our reference frame. In other words, for all we know, consciousness really is a medium outside of our laws of physics that generates free choices ex nihilo in a way that appears deterministic from a "physical" frame-of-reference.
Second, the model helps to explain the mind-body problem. Observers trapped in the kind of simulation I am describing would be convinced – just as many of us are – that there is something about their subjective point-of-view that cannot be captured in the physics of their world. And they would be right. After all, the hardware upon which the simulation is running – the processing apparatus (viz. DVD laser apparatus/processor) – would comprise their subjective point-of-view but otherwise be inaccessible to them. From their reference frame, everything besides their subjective point-of-view would be physical information (i.e. the 1’s and 0’s that make the information of their physical world). Thus, the model explains the mind-body problem. If the model is right, our reality really is made of two fundamentally different things: "hardware" (i.e. consciousness) and "software" (i.e. physical information) interacting.
Next, in the paper I argue that the P2P Hypothesis provides a better interpretation of quantum mechanics than more traditional interpretations (e.g. the Copenhagen, Everett, and Bohmian interpretations) because, unlike those interpretations -- which simply interpret quantum phenomena -- the P2P hypothesis promises to explain why quantum phenomena exist in the first place. Further, in work presently in draft I argue that P2P simulations replicate and provide a unified explanation of several otherwise unexplained features of the quantum world -- specifically:
- Quantum superposition
- Quantum indeterminacy
- The quantum measurement problem
- Wave-function duality
- “Wave-function collapse”, and
- Quantum entanglement
Physicists have explained how these features of our world work, but not why they exist. I believe the P2P Simulation hypothesis promises to provide a unified explanation of their nature and existence.
Here, in brief, is how. A P2P simulation, again, is simply one in which no individual computer serves as the “server” representing where all objects and properties in the simulation are (for the various computers on the server). Rather, in a peer-to-peer network, "the" simulated environment simply is the multitude of computers all networked together, with each having its own representation of the observed reality, and in which a netcode is used to ensure that the computers do not get too far out of line (so that they all nearly agree on where objects and properties are in the simulation). Now, here are the really neat things about P2P simulations:
- A peer-to-peer simulation just is a superposition of different parallel representations of the simulated environment on different computers on the network (viz. each computer has its own ever-so-slightly different representation of where things in the simulation are, such that the union of the different representations of "reality" is a giant superposition of alternate states),
- "The" location of any object or property in a P2P simulation is therefore also indeterminate, given that each computer on the network has its own representation of where "the" object or property is, and there is no dedicated server on the network to represent where the object or property "really" is (any object or property "really" is represented at many different positions on the network, thanks to slightly different representations on many computers all operating in parallel),
- Any measurement taken by any single measurement device a P2P network also thereby affects the network as a whole (since what one computer measures will affect what other computers on the network are likely to measure at any given instant), giving rise to a massive measurement problem (one can only measure an object is on the network by disturbing the entire network, thereby altering where other computers on the network will represent the particle as being),
- Because different machines on the network represent the same object in slightly different positions at any given instant (with some number n of machines representing a given object at position P, some other number n* of machines representing a given object at position P*, etc.) a dynamical description of where a given object/property probably is in the environment will have features of a wave (viz. an amplitude equivalent to the number of computers representing the object at a given instant, and wavelength equivalent to dynamical change of how many computers represent the object at a given point at the next instant), while at the same time
- Any particular measurement on any particular computer will result in the observation of the object as located at a specific point (thus embodying a wave-particle duality), such that
- Any particular measurement on any particular computer will result in the appearance of a “collapse” of the wave-like dynamics of the simulation into a single, determinate measurement (thus modeling wave-function collapse),
- Finally, it is a natural result of a peer-to-peer network that single objects can “split in two”, becoming entangled (in a peer-to-peer network multiple computers can, in a manner of speaking, get slightly out of phase, with one or more computers on the network coding for the particle passing through a boundary, while one or more other computers on the network coding for the particle to bounce backwards – in which case, if the coding is right, all of the computers on the network will treat the “two” resulting objects as simply later continuants of what was previously a single object).
If all this is true (a big if, I realize!), then quantum phenomena may no longer be as mysterious as they have long seemed. Rather, quantum phenomena just are evidence that our reality is structurally identical to a peer-to-peer simulation ("coded" in a system in a higher frame-of-reference inaccessible to us within our frame of reference).
Finally, there are several tantalizing new lines of evidence in favor of the theory. First, 2013 National Medal of Science award-winning physicist James Gates has found the existence of binary-error-correcting block codes embedded in the equations of string theory: error-correcting codes just like those computer programmers actually use to prevent inconsistencies in online simulations. Second, the kind of quantum feedback mechanism I predicted on p. 41 of “A New Theory of Free Will” – a mechanism for primitive, libertarian free will to interfere with and alter the normal quantum wave-function (which was also predicted far earlier by the ORCH-OR theory of consciousness) -- appears to have been observed.
Anyway, the theories -- Libertarian Compatibilism and the P2P Hypothesis -- may both turn out to be wrong. However, these are the reasons why I think, at present, they both deserve to be taken seriously.