Could we tell if … we exist only in an alien's computer?

Could we tell if … we exist only in an alien's computer?

We are probably not Sims Dr John Mitchell [email protected] Do we exist only in an aliens computer? Do we exist only in an aliens computer?

This bizarre idea seems like the stuff of movie plots, but it was popularised in more serious circles by Swedish Oxford philosopher Nick Bostrom.1 1. Bostrom N (2003) Are We Living in a Computer Simulation? Philosophical Quarterly 53 (211):243-255 Do we exist only in an aliens computer?

The Simulation Argument concerns advanced (posthuman) societies running simulations which contain conscious entities (ancestor simulations). Structure of Argument Assumption 1

Assumption 2 OR Possibility 1 OR Possibility 2

Possibility 3 Structure of Argument IF (ASSUMPTION1 AND ASSUMPTION2) THEN (POSSIBILITY1 OR POSSIBILITY2 OR POSSIBILITY 3) Possibility 1

Our species, or by a extension species like ours, is unlikely to survive long enough to reach a posthuman stage of highly advanced technology. Possibility 1

Existential risk from destructive war, environmental catastrophe, hostile AI takeover, asteroid strikes, gamma ray bursts Possibility 2 Species that do reach such a stage are unlikely to run a large number of simulations

of societal evolution containing conscious simulated characters Possibility 2 Either for ethical reasons or for lack of interest.

Possibility 3 We are likely to be living in such a simulation. Possibility 3 A probabilistic argument, based on the idea that it would be easy for such a

civilisation to run very many simulations. Thus there would be many more conscious entities within simulations than in base reality. Possibility 3 A probabilistic argument, based on the idea that it would be easy for such a civilisation to run very many simulations. Thus there would be many more conscious entities within simulations than in base reality.

We will examine this argument critically later in this talk. Structure of Argument Assumption 1 Assumption 2 OR

OR Assumption 1 Substrate independence. Assumption 1

Substrate independence; the full experience of human consciousness could be faithfully reproduced by simulation within a computing device. Assumption 2 Sufficient computing power. Assumption 2

Sufficient computing power; a posthuman society would have the computing capacity to run many such simulations at a contextually modest resource cost. Structure of Argument OR

OR Elon Musk Elon Musk https://youtu.be/J0KHiiTtt4w

So given we're clearly on a trajectory to have games that are indistinguishable from reality and those games could be played on a set top box or on a PC or whatever and there would be billions of such computers or set top boxes, it would seem to follow the odds we're in base reality is one in billions. Tell me what's wrong with that argument. Is there a flaw in that argument? Tell me what's wrong with that argument.

Lets look at the assumptions: Substrate independence is a matter for discussion and debate in neuroscience, theology, biology, philosophy we will not address it further here. Tell me what's wrong with that argument. Lets look at the assumptions:

Sufficiency of computer power for simulating an entire universe seems like a big problem. Computation has both theoretical and presumably practical limits. Ringel & Kovrizhin2 showed that its very hard to simulate large quantum systems on classical computers, implying that simulating a universe would require a quantum computer.

2. Ringel Z, Kovrizhin DL (2017) Quantized gravitational responses, the sign problem, and quantum complexity. Science Advances, 3(9):e1701758 Overflow Error? Science attempts to describe (the observable portion of) our universe.

Could a simulation model it fully? Overflow Error? Volume of the observable universe is estimated3 as 3.651080 m3 What range of volumes is feasible for an alien machine?

Orac is ~10-1 m3. 3. Egan CA, Lineweaver CH (2010) A Larger Estimate of the Entropy of the Universe, The Astrophysical Journal 710 (2):1825-1834 Overflow Error? A single computer is very much smaller than the universe that

contains it, and hence holds very much less information. ~1080 m3 ~10-1 m3 What range of volumes is

feasible for an alien machine? Overflow Error? A single computer is very much smaller than the universe that contains it, and hence holds very much less information. What possibilities remain?

1. 2. 3. 4. Simulated universe is much smaller than base universe. Only a small region of universe is fully simulated. Simulated universe has less detailed physics or coarser granularity.

Only brains are fully simulated. Hidden Assumptions The physics in their world is like the physics in ours. Bostrom1 says: The physics in the universe where the computer is situated that is running the simulation may or may not resemble the physics of the world that we observe. However, he actually addresses the question of whether a future version of humankind could simulate a universe containing conscious beings within a computing device based in our own

universe. In principle, one could imagine the simulators world having whatever magical properties one chose, but this places the simulation hypothesis beyond rational discussion.

Hidden Assumptions The physics in their world is like the physics in ours. Bostrom1 says: The physics in the universe where the computer is situated that is running the simulation may or may not resemble the physics of the world that we observe. Weve already seen that the simulating and simulated universes cant be truly alike due to the difference in volume and information content.

Hidden Assumptions A probability of living in base reality can be estimated by summing over both observers known to exist now and hypothetical observers who may or may not exist at some time in the future. The odds we're in base reality is one in billions comes from this kind of calculation. The assumption that known and hypothetical observers now and in the future can be summed

over is open to criticism.4,5 4. Franceschi P (2012) On the Disanalogy in the Simulation Argument, philsci archive. 5. Eckhardt W (2013) The Simulation Argument., Paradoxes in Probability Theory, Springer Briefs in Philosophy. Hidden Assumptions The only relevant possibilities are base reality and simulated reality.

The odds we're in base reality is one in billions comes from considering just the two possibilities: base reality and simulation. What if our universe is just one of many? What if our universe is just one of many?

Such a multiverse approach might not strictly affect the logic of the simulation argument, but it would reduce the need for a complex explanation of the apparent unlikeliness of our world and the ostensible fine-tuning of the physical constants. What if aliens could create real universes as easily as simulations? What if aliens could create real universes as easily as simulations?

Why would they bother simulating universes if they could easily create and observe real ones? Hidden Assumptions Occams razor doesnt matter. Is the simulation hypothesis unnecessarily complex?

Many people consider simulation to be a contrived and a priori implausible, explanation for our world. This is not adequately addressed in the simulation argument. Is the simulation argument any better than the dream argument? The idea that life might be (my/your/our/someone

elses) dream is ancient and was discussed by philosophers like Plato, Aristotle, Vasubandhu & Descartes. Locke & Hobbes argued against it. Simulism may be seen as the we probably cant tell if the world is real argument for the modern age. We (the indivisible divinity that works in us) have dreamed the world. We have dreamed it resistant, mysterious, visible, ubiquitous in space and firm in time, but we have allowed slight,

and eternal, bits of the irrational to form part of its architecture so as to know that it is false. Jorge Luis Borges, 1932. Is the simulation argument Intelligent Design for humanists? Simulism is a theory that posits hyperintelligent and powerful creators, but ascribes to them no moral characteristics or commandments. It argues from

observation of our world that these creators are very likely to exist. Could we tell if the Simulation Hypothesis were true or false? According to Karl Popper, an hypothesis is only scientific if it is testable. This means that we can use experimental observation and reasoning to decide whether the hypothesis is true or not. In particular, a scientific hypothesis should be falsifiable. That is, it must

make predictions which could in principle be found to be untrue, and if so the hypothesis would be rejected. Does the Simulation hypothesis make testable predictions? Would the observable world look different with or without its being simulated? Karl Popper (1902-1994) Could we tell if the Simulation Hypothesis is true or false?

Our simulations typically use directional grids and discrete time steps. We might (speculatively) expect to find that distance and time were quantised and that space had directional properties (anisotropy). Theres currently no good evidence that distance and time are quantised, even down to the Planck scale.

Could we tell if the Simulation Hypothesis is true or false? Simulators could run a simulation with insufficient physics to describe its universes origin, if they started running from t > 0. If we lived in such a simulation, our cosmologists might be endlessly puzzled.

Could we tell if the Simulation Hypothesis is true or false? If we lived in a simulation and our physics were like their physics, we could expect to be able to develop effective quantum computing and conscious, but not overly threatening, AI. We might also expect that other existential

threats like nuclear war and environmental catastrophe were avoidable. Could we tell if the Simulation Hypothesis is true or false? Failure to find properties (naively?) expected of a simulated world might make simulism seem less likely. However, its unlikely that we could tell for

sure either way. Tell me what's wrong with that argument. 1. Simulation argument makes optimistic assumptions about sufficiency of computing resources. 2. Simulation of a whole universe requires simulators physics to be both like ours, so that the scenario can be rationally evaluated, and unlike ours, because a computer is very much

smaller than a universe. 3. Simulation argument ignores all other existential possibilities besides a base reality universe and a simulation. 4. Occams Razor suggests that the simulation hypothesis is unnecessarily complicated.

Recently Viewed Presentations

  • Chapter 8

    Chapter 8

    Chapter 11. Quiz Questions. ... An athlete has sustained a sprain to the cervical region and is having radiating paresthesia down the left arm. What should be done for this athlete? ... In a brachial plexus injury, pain is usually...
  • WIN-05 Talk - Sinica

    WIN-05 Talk - Sinica

    (nN) T < 1 keV Results : mn(ne) T ~ 1-100 keV On-Going Data Taking & Analysis SM s(ne) T > 2 MeV Reactor Neutrino Interaction Cross-Sections mass quality Detector requirements 1 counts / kg-keV-day Magnetic Moment Searches : Highlights...
  • 55:148 Digital Image Processing Chapter 11 3D Vision,

    55:148 Digital Image Processing Chapter 11 3D Vision,

    The fundamental matrix. Relative motion of the camera; the essential matrix. Estimation of a fundamental matrix from image point correspondences. Camera Image rectification. Applications of the epipolar geometry in vision. Three and more cameras.
  • NATIONALISM - Central Bucks School District

    NATIONALISM - Central Bucks School District

    Rising Nationalism. Nationalism was the most powerful . ideal of the late 1700s/ early 1800s. Increasing numbers of peoples . desired self-government and control. over their own lives. Inspired by Enlightenment ideals . and by the American and French ....
  • OBJECTS IN OTHER CLASSES:  Create a class for

    OBJECTS IN OTHER CLASSES: Create a class for

    Composition, Inheritance, and polymorphism. Composition: defining a new class that is composed of other classes. Student class was composed of the Course class. Deck class was composed of the Card class
  • LDAP Content Synchronization - OpenLDAP

    LDAP Content Synchronization - OpenLDAP

    LDAP Content Synchronization Kurt D. Zeilenga Jong Hyuk Choi OpenLDAP Project IBM Research ... DNs are not stable identifiers! ... If the server can reliably determine which entries previously present in the shadow copy which are no longer present in...
  • Chapter 5.2 notes: Hess&#x27; Law chapter 5.3 notes: bond enthalpies

    Chapter 5.2 notes: Hess' Law chapter 5.3 notes: bond enthalpies

    Enthalpy cycles. Sometimes it's hard to determine the enthalpy change directly, so an indirect route is necessary. This is shown in the example cycles
  • Thermodynamics - FHI

    Thermodynamics - FHI

    Langmuir Hinshelwood Mechanism Both species are adsorbed, adsorption follows Langmuir isotherm (see class next week) Eley Rideal Mechanism How to Derive a Rate Equation I How to Derive a Rate Equation II How to Derive a Rate Equation III Structure...