Information theory clarifies why a physical world can’t output itself, because that would require a context specified in advance based on physical states that are not yet defined. As McCabe concludes:
“All our digital simulations need an interpretive context to define what represents what. All these contexts derive from the physical world. Hence the physical world cannot also be the output of such a simulation.” (McCabe, 2005).
The same physical world can’t be both processor and output but can one part save and reload another? Could we save and reload physical reality as we do information? To save a matter scene as we do a game scene requires a data structure that existed before the scene did, so it can’t be based on matter. To save a game requires a pre-defined data structure, so if a new game version changes that, old saves don’t work! To reload a universe would likewise require a known data structure else the result would be nonsense, just as reloading a WordPerfect file into Word gives gibberish because Word doesn’t know WordPerfect’s “interpretive context”, as McCabe argues.
Imagine our universe frozen in a static state at a moment in time, as innumerable physical states that have no information at all in themselves. Information requires an observer to decode it but who could “read” it? Not us, as we would be frozen too! A frozen world without an observer would be as empty of information as this page is without a reader. To save and reload a physical universe one must define its data structure and exist outside it!
When we store movies in file to replay later, we need an agreed format like mp3. To play the file, we also need something to dynamically play it, like a laptop or projector, so without a projector to play it, there is no movie. Likewise, a computer plugged into a power source is needed to run a program file. So even if one could save the universe as static data, what “projector” using what power source could reload and run it? Information stored as a static file always needs a dynamic system to reload into.
Trans-humanists take the mind to exist entirely as information encoded in the physical brain, so expect that we will soon be able to upload the mind and download it to a new, younger, body and so live forever. However even if we could make a perfect physical copy of a brain at a moment, that is no better than taking a photo in a movie theater and taking a photo of a movie doesn’t “store” it, as one frame is not a movie. Even many photos taken and replayed in sequence doesn’t “resurrect” the person, any more than playing a movie of a dead person resurrects them. Recording a physical scene and replaying it as a hologram may emulate a dead relative but it doesn’t recreate them. The viewing experience would be real but the “person” viewed wouldn’t experience life as an observer. Information based on physical states still needs an observer, so one has to copy the observer as well. In the same way, saving a multi-player online game doesn’t save the other people playing it because they aren’t “in” the game. Chapter 6 returns to the issue of the need for an observer in more detail.
Since genes are information, why not copy genes to create a biological copy of ourselves? This can be done but nature already does it, as identical twins are from copies of the same original cell. Yet the result is twins that are two different people not the same person. A clone of me doesn’t make another me but another person. Even if that copy of “me” has experiences, it still isn’t me if I don’t experience what the clone does.
Something is missing in the reloading reality idea and it is the quantum reality behind physical events. If that is what “runs” physical reality, the question reduces to whether quantum processing can be saved and reloaded.