Comments on: Singularity Summit 2011

By: Emil

Emil — Sun, 25 Sep 2011 20:12:59 +0000

#6, I dont about urs, but my computer does hav lots of GHz mor than a few years ago. Ofc, the trend is just to add mor and mor cores insted of significantly increasing the speed of the one’s. If we just add them up, my CPU is 3.3*4=13.2 GHz; add to that the speed of the in-bilt GPU unit. According to Intel, it is 850 Mhz. Adding gives rofly 14 GHz then. Pretty close to ur 15-20 GHz. The CPU also has a turbo feature, at a max of 3.7 GHz and so does the GPU at 1.1 GHz. Adding those gives a number in the 15-20 range. And then ther is the fact that my CPU is not the fastest on the market AFAIK.

By: mustapha mond

mustapha mond — Fri, 23 Sep 2011 22:54:05 +0000

Will they be selling vitamin supplements at the summit too?

By: Clark

Clark — Fri, 23 Sep 2011 15:10:42 +0000

Bob, Moore’s law just deals with the rate of increase of transistors not frequencies of internal clocks. What happened years ago once they started reaching certain limits on chip design due to high frequencies and some electron tunneling problems was to simply go off in different directions. (Primarily parallel computing – although software hasn’t caught up there as well so the perception of power is off since most people use single threaded software) More interesting (no pun intended) is the recent companion to Moore’s law, Koomey’s Law, that deals with power usage. However Moore’s law still appears to fit the data pretty well and is only indirectly about computing power. (Although it deals with that well also)

By: Clark

Clark — Fri, 23 Sep 2011 15:06:52 +0000

I think that Watson media spotlight (i.e. extended ad for IBM) had everyone donating the winnings to charity. Jennings is actually pretty clever and has been doing the rounds for his new book on maps. It sounds pretty funny and interesting at the same time. A difficult trick to pull off on topics like that. (Here's a review although it's an LDS blog site - Jennings often guests at various major LDS blogs) I'm still not sure what the Watson event proves. I'm much more interested in seeing how IBM's efforts to leverage the technology into the medical field turns out. I hear there is somewhat expected opposition from some doctors. Yet that really does seem like a field Watson would excel at. I'm really not at all impressed with most doctor's ability to diagnose problems. I've not been able to find a lot of reports about how Watson in medicine is actually proceeding beyond comments here or there at various blogs. What does get written up seems a tad too much influenced by the IBM PR machine.

By: bob sykes

bob sykes — Fri, 23 Sep 2011 11:18:13 +0000

Moore’s Law ceased operation several years ago, otherwise we’d have 15-20 GHz personal computers instead of 3 GHz models. Perhaps IBM’s 3D chip will revive it. However, the singularity remains nonsense. The people who go off to conferences to talk about it are a bunch of Jeremy Rifkins.

Several years ago, Hubert Dreyfus (“What Computers Still Can’t Do,” MIT, 1994) described the state of the art of AI. His conclusions are still valid. We are now going on 40 years of failure and deceit. Watson’s success on Jeopardy (I watched it with my wife) depended entirely on response speed. It operated well within the response time of the humans and go there first. The accuracy of its guesses was no better that the humans. That, of course is a considerable programming achievement, but it points out the enormous computing potential of the human mind. And notice how small the human brain is compared to any computer.

One fact that continually escapes computer people is that the human brain is a collection of analog machines each dedicated to a few related tasks. Trying to emulate this with a general purpose digital machine is doomed to failure. Any possible digital machine is necessarily slower than an analog machine. The inferiority of Watson compared to its human competitors is striking. A huge, slow digital machine that can only solve one problem against a human brain built on wetware (electrical impulse speed is what? 1 m/s?), with many specialized modules, that can walk off the set and lead a life.

By: Justin Giancola

Justin Giancola — Fri, 23 Sep 2011 09:11:51 +0000

This is where people who want to be single for life meet up, right? ho ho ho ;p
…and then proceed to annoy each other to death, ala reductio ad highlanderum.

I love how they say Watson won a million dollars. I wonder what “Watson” will be spending “his” – if it is a male name shall I refer to it with male pronouns?- money on. It would be nice if they dropped a mention did the designers keep the money; did it go to a cause?

btw, anybody else ever jam with one of their professors? (did this today) by this I mean played music.

By: Razib Khan

Razib Khan — Fri, 23 Sep 2011 07:36:48 +0000

kurzweil is the media front. there’s plenty of skepticism in the audience about any given claim in my experience.

By: rick

rick — Fri, 23 Sep 2011 07:15:09 +0000

These always struck me as a big circle jerk of charlatans who are as far from being experts on the things they’re talking about as the man on the street. For instance, Ray Kurzweil appears to know nothing at all about actual computer science research. I take it the “gerontologists” aren’t exactly doing much more than giving a summary of research interpreted through a very optimistic lens.

By: Razib Khan

Razib Khan — Fri, 23 Sep 2011 05:25:31 +0000

#2, though some bloggers of prominence can get media passes.

By: Neil

Neil — Fri, 23 Sep 2011 04:46:10 +0000

Drat, wrong coast.