Rich and I got in a big discussion about this over lunch the other day and I wanted to see if it could spur some debate in the vast world spanning techRivet Community.

It goes a bit like this.

Assumptions:

1. Hard AI is eventually possible. [Rich BTW does not agree that it is]
2. Moore's law holds across technology implementations.  Meaning it is not tied to just silicon based semiconductors but will continue its pace of increasing speed as replacement technologies are developed.

If those assumptions are true then it follows that:

1. Hard AI is inevitable.
2. At some  point Timmy the 14 year old nerd will have a simulated universe in his mom's basement on his $400 Dell.  Complete with trillions of legitimate sentient beings.
3. We have no way of knowing if we are not already on Timmy's machine.

If those assumptions do not hold then:

1. Is that an implication that there is something unknowable about the human mind? 
2. A soul perhaps?

Of course you can take the practical way out and agree with Descartes "I think therefore I am." and call a spade a spade and that this whole argument is just foolish sophistry.

...but what's the fun in that?