If there is an unknown process in reality that does this then why can't a computer be designed to take advantage of this once we know that. The burden of proof is on you to prove the process not hand wave it away.
Any mechanism that could make the human mind non-algorithmic (none have been discovered so far) could still be harnesses by synthetic life forms. Penrose's non-algorithmic mind idea has far fewer ethical or religious implications than a lot of people seem to think. In his world thinking machines are still possible, the only trick is that Turing machines can't do it.
Personally, I don't think that there is a special mechanism (or pixie-dust) that separates my brain function from that of my cat. Of course, there are some qualitative differences which I clearly don't understand. But, blaming the fact that I don't understand the differences on the lack of some scientific breakthrough doesn't seem to clear the whole problem up : It's passing the buck.
IMHO, when the 'trick' to consciousness is found, people will look back and see that what seemed like pixie-dust was just a change in Order() of the system as a brain gets more complex. And there's nothing to suggest that there couldn't be another similar 'upshift' in consciousness beyond that which regulation-sized human brains can do...
This is a common misconception about Penrose for some reason. I think what he's really saying is that there is some as-of-yet-unknown physical process that is responsible for consciousness. In order for us to discover what exactly that physical process is, however, we will probably (though not necessarily) have to completely revise our understanding of quantum mechanics. For example, we might need a theory that successfully integrates QM with relativity.
Wikipedia has a decent article on hypercomputing(http://en.wikipedia.org/wiki/Hypercomputation#Hypercomputer_...) (i.e. computers that can do more than Turing Machines), which is essentially what Penrose believes is responsible for human intelligence. Many of these weird hypercomputers use non-quantum properties of the universe to achieve hypercomputation. Two common examples are creating and collapsing different threads in space time, or computing with arbitrary precision real numbers.