Tuesday, January 13, 2009

Smart Grid

It seems that the new initiative to create a huge smart grid is very good. This could be the start of a smart world. Real time tracking of everything. the ability to , in real time see the impact of such things as financial stimulus, health care etc.
the minimization of waste would be enormous. Naturally supercomputing sits at the heart of it all. Negative and positive feedback loos with evolutionary programming continually fine tuning the smart sub system.
It would be good if my house knew when to buy power. Now if only those guys in Texas with the great new battery would start selling them.
Time to get involved and go to Austin for a kick of think tank. Should be fun

Monday, December 29, 2008

The singularity seems to be going double exponential

A lot of people seem to be talking about the 1k genome during the 2009 early 2010 time frame. This would be several years ahead of what conventional singularitists would have thought. We have 10 ^15 supercomputers running conventional benchmarks and Gpu systems about to run about the same for 1/10 the price.
It is now becoming clear that having the first such system would enable several possibilities that are , by today's standards not achievable. Drones with near human ( 20 percent) of human capacity, instant online content created in photo realistic ways in real time. This will be good because i want to watch more of a certain types of movies. they just cost to much with humans involved.
what will it be like to simulate a real cell or to create a model of functioning microchondria.. could we see de grays ( http://en.wikipedia.org/wiki/Aubrey_de_Grey) work come much sooner than thought possible.
i feel it in the wind so to speak. things are changing faster than i thought.
I dont think we have a way to fund the start ups. They will not need so much money ( 250 to 500 k) to do there "show" me. They will model and or make things go very fast
They wont need studios or human trials of huge factories to prove there point and create value. the question is, can the common man/woman put in micro investments so we can by pass old style vc structures that require that projects be large.
Do they care that 500 k turns into 5 million, instead of needing 20 million to just get in a game.
i do think its time to bring in the normal day to day person. We have just been shown that traditional investment advice is not good. They thought they were taking little risk...lol..wrong. time to put a tiny bit in short term de risking projects.
any one want a petaflop for 2 to 5 million
Peace
Tim

Sunday, November 30, 2008

mice mice and more mice

Yet another great sucess for some therapy that works in mice. who cares anymore. They never seem to get to humans. none of the test subject have had there genome sequenced. Its time to focus more on how humans react reather than mice that are proclaimed to be so close to us in genetic terms, yet so far when it comes to the real world. Yest , its true we are 1 year to 18 months away from the full sub 1000 dollar genome. But really we need a 100 dollar target. Imagine having to compare all the data in a full multi dimensional crossplot, and to do it on each trial patient. Hech even back of a step and do it on all the mice. Hundreds of billions out the window in th eruch for the next super cure.
time to back of and build the tools first.
In the mean time it might be best to move to spain , drin a lot of red wine , while building a box that can do multi human smith waterman sequence matches in 10 minutes.
Only one small problem with that.. needs 20 million ( in parts ) not wine.




http://www.physorg.com/news147272893.html