Since at least the reader of this text exists, something exists. Isn’t that great? Now, why do things exist? It must be costly to create things out of nothing. Maybe that’s okay because something has always existed. Here are some ideas relating to the question of why anything exists.

This one is listed only because it’s popular. Obviously, it doesn’t help us at all because it doesn’t explain why a creator exists in the first place.

Here, we consider that the information contained in the world grows over time because the world is the product of a random data generator that underlies everything. This theory assumes some fundamental notion of time and an origin of time where the world starts from nothing and gets more complex as random data or “matter” gets produced. This theory feels satisfying when we consider the existence of an initial state containing zero information. It is dubious for other reasons:

- Like for the god hypothesis, it doesn’t explain why a random data generator would exist in the first place.
- This model assumes time and the origin of time. This would have to be modeled as part of the random data generator, making it non-trivial.

If everything exists, then the world taken as a whole is trivial and it’s equivalent to not existing at all. A difficulty would be to come up with a satisfying, rigorous definition of “everything”. Everyone with perfect knowledge would have to agree with that definition since the theory rests on there being a single, natural everything.

We could have a new “theory of information” where information held by a model collapses as the knowledge about this system approaches completeness.

We could model a system using a collection of variables representing unknowns and constraints between these variables. The constraints form the structure of the model. The size of a maximally-compressed structure of the model gives us its information content. The complexity of a correct but partial model relates to the number of possible instantiations of variables in the model. If two variables are constrained to always have the same value, they’re considered redundant and collapsed into one. This leads to the elimination of all the redundant variables and of all the variables constrained to take a single value. A complete world model would have only one possible instantiation of variables, making it as simple as the empty model containing no variable and no structure.

Maybe we need a hypercomputer and a hypermind to make sense of why anything exists. A hypercomputer is a hypothetical computer that can carry out an infinite, countable number of operations within a finite number of steps. A hypermind would be a mind that could emulate a hypercomputer if it feels like it.

*Martin Jambon, November 27, 2022*