[extropy-chat] D-Wave premiere of 16 qubit processor

[Ben Goertzel]

John K Clark wrote:
> "Damien Broderick" <thespike at satx.rr.com>
>
>   
>> They're still claiming that < Quantum-computer technology can solve
>> what is known as "NP-complete" problems > although critics have shown
>> that this is an excessive claim for their own system.
>>     
>
> If what they say is true then this is HUGE, far bigger than cold fusion
> would be even if the damn thing worked, which it doesn't. However I have a
> sneaky hunch the company is exaggerating their progress in this area just a
> tinny tiny bit, but I sincerely hope I'm wrong. If I am wrong then the 
> Singularity will happen much much sooner than anybody expected,
>   

An effective, scalable specialized quantum-computing-based 
problem-solver would certainly be really exciting,
and that is what these guys claim to have constructed.

However, it is a fairly long way from a powerful tool like that to a 
Singularity ;-)

The most direct way I can see a tool like that leading to Singularity is 
via an AGI being implemented on it
and I can tell you that a machine like that definitely does NOT solve 
all the problems of AGI.  Though it
certainly would help.

For instance, if we had one of their 1000-qubit boxes to play with, then 
that would cut down the amount
of hardware needed to run a mature Novamente system by a couple orders 
of magnitude ... and would
probably save us a dozen man-years of work in accelerating our 
evolutionary procedure learning
and data mining components via clever algorithm tweaking and distributed 
processing.  OTOH, it would
possibly take at least an equivalent amount of work to port our 
algorithms and design to the QC in
an effective way ;-)

The history of parallel computing shows that porting algorithms to novel 
computing infrastructures can
be trickier than it first appears...

-- Ben G

> even Eliezer.
>
> Hmm, now that I think about it, maybe that's a little too soon; the 
> Singularity will be a meat grinder and I'm unlikely to come out 
> of it in one piece. No matter, it's probably just a PR blurb to boost 
> the company. Probably.
>
>   John K Clark
>
>
>
>
>
>
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
>