Flexibility in Grover's Algorithm?

[.Scott]

Grover's Algorithm is a QM process for finding a high-scoring element in a large data base.
(http://arxiv.org/abs/quantph/0010040)

My specific curiosity if of a biological system that would generate a superpositioning of data that represented potential intentions, to estimate the consequences of these potential intentions based on previous experience, then to rate these consequences according to the likely "goodness" to the biological system, and then to pick out this "best of intentions" using Grover's Algorithm.

Although I do not doubt that some QM information processing can be done in the "wet" environment of a biological system (http://phys.org/news184423418.html), the "averaging" step in the algorithm impresses me as potentially challenging for such wetness - especially when it needs to be repeated.

So here are my questions:

If the Grover Algorithm is implemented in a looser fashion, would the results fall apart completely, or could you get a "good" intention selected - even if it may seldom be the "best".

The algae photosynthesis demonstrated that biological molecules can be used for processing superpositioned states to the advantage of the organism in a wet environment. Making use of such devices in a brain for the purpose I described above would require superpositioning that stretched across many such molecules and would involve many qubits (or QM analog equivalents) involved in a dingle superpositioning. Given the technology for creating a brain at all, are there any theoretical or practical restrains to keep this from being implemented?

Thanks,

Scott Bowden

 

[Greg Bernhardt]

@.Scott did you find any more insight on this topic?

 

[.Scott]

Certainly over the past 6 years many have gotten a better feel for what happens when your QM system is not reliably isolated. I've played with IBMs online Quantum computer server. They run your program hundreds of times and give the statistics as your result. Any "intention generator" is going to need a follow-on step where the quality of the intention can be more thoroughly reviewed. So an implementation of the Grover algorithm that is noisy could be tolerated.

So the answer to the original question is "Yes, but...".
The "but" is that we have no idea what this wet mechanism is, what its error rate is, what is its bandwidth, etc.
A QM System based on Grover's Algorithm that performs nearly as well IBMs online QM computer would be great - but it would be much bigger than what IBM has demonstrated.

I can ask if a biological QM system could, in theory, be useful. But really, the question only has teeth if we have a case-in-point.