Follow Slashdot stories on Twitter


Forgot your password?

Comment Re:Give it time (Score 2) 114

"Geoff Hinton knows how the brain works"

IBM's Watson doesn't understand anything. It just measures statistical correlations between pairs of things. Same for Siri

IBM's chips are based on spiking neurons. These will let existing algorithms run extremely fast, and are very useful, but are not "how the brain works".

There are fundamental (large) gaps in our understanding of how we think. We don't really know how it works at all, despite the hype.

Comment Not the only game in town (Score 2) 364

There are alternative theories, and they probably ought to get much more attention than they do. I think the fact that science is a career for people these days, makes them more keen to play things safe.

For example, Professor Mike Mcculloch's MiHsC seems to predict a variety of otherwise anomalous observations rather well, without endless fudge factors. He's a respected academic, but seems to get little mainstream scientific attention.

Comment Models (Score 1) 525

I believe in science, but I do wonder about the accuracy of climate scientist's models. I remember reading Richard Feynman's account of a scientist's rat-in-a-maze experiment - "STILL, the rats could tell". It detailed how hard it is to control all the variables in an experiment - in that case, to make sure the rats were operating under the rules you thought they were. The same principles would surely apply to climate models, where the system they're modelling is enormously more complex.

Comment Re: Marijuana's capacity to REVEAL TRUTH (Score 1) 291

Nuclear weapons have killed a little under 250,000 people world wide since 1945, whereas traffic deaths in 2010 alone were over a million worldwide. Therefore cars are more dangerous than nuclear weapons.

Such comparisons aren't very useful, because they ignore the fact that the weapons (thankfully) aren't used often.

Comment Local minima (Score 1) 249

Their model isn't necessarily inadequate. Perhaps the cooperative strategy was simply easier to arrive at through evolution. The extortion strategy might be in a hard-to-reach part of behavioural state space. It's taken these brilliant mathematicians a good while to find it, after all. If evolution finds a suboptimal, but still beneficial strategy, it can be hard to subsequently jump out of that local minima to reach an even better solution.

Comment Re:Slave Labour is certainly profitable (Score 2) 534

I'm using a PC, it was assembled in the UK, but as you point out, the components were surely made in the far East.
So what? Is your point that because we all use computers, and wear clothes that were made by workers in terrible conditions, that it is wrong to criticise those conditions?

I never said Apple were the only company that does this, but they are the biggest, and they have the largest gulf between their polished, doing-good-for-all image, and the reality.

"Falling in love makes smoking pot all day look like the ultimate in restraint." -- Dave Sim, author of Cerebrus.