Notes on two presentations by Gordon Bell ca. 1998
|
From: nmm1 on 6 Feb 2010 04:27 In article <4B6CDCA2.9090107(a)patten-glew.net>, Andy \"Krazy\" Glew <ag-news(a)patten-glew.net> wrote: > >> well know older competitor to Cambridge. > >Stamford? :-) > >(Not Stanford. M, not N. > >http://en.wikipedia.org/wiki/Stamford,_Lincolnshire#Education Nice one! I hadn't heard of that, and should have done. It's more recent than the place I work in, of course, but .... Regards, Nick Maclaren.
From: Bernd Paysan on 6 Feb 2010 17:56 Robert Myers wrote: > Anyone who used the old machines can remember how much science could > be done with them. Can we now do billions of times more science? Obviously depends on the big O() of the problem. If your problem is O(n³), then you can do a thousand times more science now ;-). -- Bernd Paysan "If you want it done right, you have to do it yourself" http://www.jwdt.com/~paysan/
From: Del Cecchi on 6 Feb 2010 19:08 Robert Myers wrote: > On Feb 4, 4:05 am, n...(a)cam.ac.uk wrote: >> In article <7sv55fFfm...(a)mid.individual.net>, >> Del Cecchi <delcecchinospamoftheno...(a)gmail.com> wrote: >> >> >> >>> You can do a billion times more science because you can do things now >>> you couldn't do at all then. Like model the weather hundreds of years >>> in the future. ok, the climate not the weather but whatever. >> One can debate how much that is due to computing power and how much >> to more and better data, but the computing power is certainly used >> for that. And might be essential for it. > > And the fact that we now how the capacity to *appear* to be able to do > this kind of modeling and that (mirabile dictu) the modeling predicts > problems that require massive funding don't make you at all > suspicious? > > As I've pointed out, prospective modeling of weather patterns a few > months in advance has a dismal track record. Given that reality, why > should I take climate "modeling," which is little more than a not > always successful exercise in retrospective curve-fitting, any more > seriously? > > And I strongly suspect that Del had his tongue in his cheek. > > Robert. > Me? A Denier? :-) But we can simulate a million gate chip and time it too. Which we couldn't do either. But my point is that ratios are meaningless if we can do things that were impossible before. Other things are still impossible. :-( Accurate Climate forecasting seems to be suffering from all four problems: lack of enough data, lack of enough understanding, a predetermined answer, and lack of computing power
From: Robert Myers on 6 Feb 2010 19:23 On Feb 6, 7:08 pm, Del Cecchi <delcecchinospamoftheno...(a)gmail.com> wrote: > Robert Myers wrote: > > On Feb 4, 4:05 am, n...(a)cam.ac.uk wrote: > >> In article <7sv55fFfm...(a)mid.individual.net>, > >> Del Cecchi <delcecchinospamoftheno...(a)gmail.com> wrote: > > >>> You can do a billion times more science because you can do things now > >>> you couldn't do at all then. Like model the weather hundreds of years > >>> in the future. ok, the climate not the weather but whatever. > >> One can debate how much that is due to computing power and how much > >> to more and better data, but the computing power is certainly used > >> for that. And might be essential for it. > > > And the fact that we now how the capacity to *appear* to be able to do > > this kind of modeling and that (mirabile dictu) the modeling predicts > > problems that require massive funding don't make you at all > > suspicious? > > > As I've pointed out, prospective modeling of weather patterns a few > > months in advance has a dismal track record. Given that reality, why > > should I take climate "modeling," which is little more than a not > > always successful exercise in retrospective curve-fitting, any more > > seriously? > > > And I strongly suspect that Del had his tongue in his cheek. > > > Robert. > > Me? A Denier? :-) > > But we can simulate a million gate chip and time it too. Which we > couldn't do either. > > But my point is that ratios are meaningless if we can do things that > were impossible before. You mean like before Claude Shannon's Master's Thesis. How do computers look if you put *that* in the denominator? > Other things are still impossible. :-( > Accurate Climate forecasting seems to be suffering from all four > problems: lack of enough data, lack of enough understanding, a > predetermined answer, and lack of computing power We don't know the state of oceans and atmosphere at t=0, but we know it with 90% certainty in fifty years. Amazing. Just absolutely amazing. The lack of data problem would be more expensive to solve than the lack of computing power problem. Robert.
From: Del Cecchi on 7 Feb 2010 00:03
Andy "Krazy" Glew wrote: > nmm1(a)cam.ac.uk wrote: >> In article <4B6CDCA2.9090107(a)patten-glew.net>, >> Andy \"Krazy\" Glew <ag-news(a)patten-glew.net> wrote: >>>>> well know older competitor to Cambridge. >>> Stamford? :-) >>> >>> (Not Stanford. M, not N. >>> >>> http://en.wikipedia.org/wiki/Stamford,_Lincolnshire#Education >> >> Nice one! I hadn't heard of that, and should have done. It's more >> recent than the place I work in, of course, but .... > > My dad, and several older males in my line, are/were Stamford old boys. > > (Not rich: just tenant farmers in the area from a family who valued > education and were smart and lucky enough to to get scholarships.) > > I often joke that my education is along the lines of Stamford in England > and Tsunami Tsate Tsuniversity in Oceanside, Oregon. And here I was wondering what fine educational institution was in Stamford, conneticut besides, was it cbs labs? |