Thu, Nov 25, 2010  Permanent link
What's your thoughts on this? 
BenRayfield     Fri, Nov 26, 2010  Permanent link
dmitri, it describes a way to run certain brain-like calculations using a million times less energy and therefore able to run more of them at a time without overheating. I expect they will learn much about brains, and if they eventually build enough of them to simulate a whole Human brain, it could learn to make all the same mistakes I described above and help us create World War 3 faster and more efficiently.

Brains are not a good software design to copy. Most of them are not general computers as proven by their inability to learn the following as intuition instead of something written on paper: If I flipped 2 coins, both landed, and at least 1 landed tails, then whats the chance both landed tails? The 2 coins can be replaced by any 2 events, the question asked in that context, and whatever learning from the correct answer of 1/3 (not 1/4 or 1/2 which most people say) is not transferred to the new context. Human brains have extreme difficulty with simple bayesian questions, a flaw which bayesian networks do not have. As I explained above, "The average person is dangerously insane, preferring to buy health insurance instead of do things to avoid their own death (and extinction)."

They're doing the wrong calculation a million times more efficiently. I expect they'll make a lot of money from it, since its the calculation those with power are doing.

I have a question for them... When they build their supercomputer of brain simulations, will it answer that simple math question with a 1/3 or with the same 1/2 and 1/4 which most people answer, and if it answers like average people do, then do we really want a supercomputer that can't answer a simple statistics question?

For the same reason, an ancient computer running a chess search-tree software with bayesian optimizations will still beat their Human simulation at the game of chess. What are they really accomplishing?
gavinkeech     Fri, Nov 26, 2010  Permanent link

focus on what is seen to be best avoided, best be avoided in mentioning.

let's build a means of utilising thought-forms to allow increased evolutionary progress to occur.

when evolutionary progress begins with the right focus and continues those ways(multiple), negative possibilities are absorbed(removed through interplay) to allow growth and thrivability.

the exaptive form/non-form of polytopia, has lots of what is happening on earth as it's emerging. all of these posts are comments to your questions and answers which are of good sight, of what to expect, if we remain on fluid pathways with the right choice of focus on perspectives.

the details : are the emotions attached to questions and answers we make, let's not attempt to form strong concepts of what is wrong, and begin to build without sight of the problems we want to avoid. they can be there as possibilities, but the discovery of the future happens through urgency, that is enough to continue what we are, and go where we can help each other enter.

     Fri, Nov 26, 2010  Permanent link
It's not like they're doing this to process complicated mathematical equations. From their page, this is what they're trying to accomplish:

A promising alternative to GPUs and FPGAs, Neurogrid will make the computational power required to explore various hypotheses about how the cortex works affordable. Neurogrid's speciality is modeling interactions between cortical areas, of which there are over three dozen in the visual system alone, connected by lateral, feedforward, and feedback projections. Feedback projections constitute about half the total—virtually every area projects back to the areas that feed it—yet their role remains mysterious. One hypothesis is that they integrate these areas' myriad representations into a coherent percept. Another hypothesis is that they mediate attention, zooming in on the most informative area and excluding the others.

We have things like the blue brain project using so much resources to understand something that in reality runs on 20watts of electricity.

This is going to allow computers to perceive as we do - Hopefully increasing the symbiotic relationship between man and machine so both become more useful. In the meantime it's not like we're going to stop using regular computers, it's just an adjunct.
BenRayfield     Fri, Nov 26, 2010  Permanent link
Yes it will be useful for those kinds of things, but its off topic here. This thread is for summarizing the most important things happening on Earth, and I defined the minimum importance this way: "Details like disagreements between countries are not important enough for this thread." You could write about patterns in advancing brain simulation technology and what that leads to, but the specifics of any 1 technology are not what this thread is for. My previous response was in the context of those patterns I wrote in the first post.
     Sun, Nov 28, 2010  Permanent link
Please reply with the summary of whats really happening on Earth, only the most important patterns, where they lead, and what we might do about it. Too many people get stuck on the details.


Maybe that IS the problem. Maybe we're so busy trying to make reductive theories of what is up, which inevitably leads to missing things out, as this problem is less of an absolute and more of an ongoing process that slips between our fingers when we try and grasp it like some kind of non-Newtonian fluid. Maybe we could do well with a model-free systematic interpretation of this problem: 
BenRayfield     Sun, Nov 28, 2010  Permanent link
I agree we should put all these patterns into software (I'll eventually build something like that), but no software exists yet thats smart enough that it would help us understand it better than simply writing the most important patterns here. I didn't have a "problem" writing what is probably half of the important patterns on Earth in the first post. I'm asking for the other half.