Some Things I'm Not Completely Sure About...

1.. P=NP is either a number theory problem (2-SAT and 3-SAT could be converted into encodings for different types of numbers?) or a physics problem (a question regarding the nature of spacetime).

2. Fractal-like systems and/or systems with very simple rules can resolve inconsistencies between quantum physics and the general theory of relativity and may be important for explaining the genesis of galaxies, Wolfram style.

3. Computing that is more "reconfigurable" (like FPGAs, evolving processor architectures, programming for many-core systems) in increasingly higher levels of abstraction will bring big results in machine intelligence.

4. The first three items are related.