October 24th, 2004


Some stats

I just popped here and went through the last year's friendings and unfriendings.

63 people have unfriending me
130 have friended me.

Leaving me 67 up.

Of course, some of the friendings are repeats - at least one person added, dropped and added me again - but the overall change is still right.

I have 200 friends, of which 163 friended me back (yup, there are 37 people out there that I read who don't read me), and 24 people I haven't friended back (largely because I already have 200 people I'm keeping track of).

I've posted 3015 entries, made 9108 comments and received 13753 comments in return. I don't know if the 13753 comments includes the ones replying to my comments elsewhere...

I also have 142 pictures.

I seem to have been busy for the last 3 years...

One for the tech-heads

This is a video of a presentation by the lead architect on the Pentium 6 on chip design and the future of CPUs. If you have any interest in the future of silicon, it's well worth a watch.

It's also an hour and a half long, but watch the first ten minutes. I did, and now I'm hooked. It's a presentation made _after_ he left, so he has some fairly scathing things to say about politics in big companies.

(The upshot so far is - we cannot carry on in the direction we are going, heat problems are just too much, complexity problems are just too much and we are running out of new big ideas to improve performance. We are, instead, relying on multiple little improvements, each of which adds a layer of complexity and means that the chips are much harder to debug).

Oh - and it's an ASX file, so Windows Media Player only. Well worth grabbing a machine with that on it though.

Don't ask, don't get

I emailed the guy behind the rat brain experiment I mentioned yesterday.  I was curious about the way that training could be done without some way of feeding back that the right or wrong outputs were being received.  I was most gratified to get a response:

The neural flight control that is being reported is very rudimentary. The in-vitro network of rat cortical neurons simply controls the pitch and roll of the aircraft to produce straight and level flight, the neural equivalent of an autopilot. This is accomplished using an effect reported by Eytan, D., Brenner, N., and Marom, S., Selective Adaptation in Networks of Cortical Neurons. Journal of Neuroscience, 2003. 23(28): p. 9349-9356 in which "high" frequency stimulations (once every second) was reported to depress the response of the network while "low" frequency stimulations resulted in an enhanced response. For our system we tied the network's response to the control surfaces, dedicating stimulations on one channel for pitch, and a second for roll control. Each channel is stimulated separately, and the response (PSTH) is recorded. Control movements are proportional to the current error from straight and level by mapping the error (0 to 180 degrees) to the interval 0 to 100 ms of the PSTH and integrating the difference in response before training, to the current or enhanced or depressed levels. The more error, the more the control surface is moved. The networks only gradually control the aircraft since the Marom effect requires over 15 minutes to develop. The two frequencies are then used to adjust these weights (i.e. number of spikes in the PSTH) to produce optimal flight. The neurons/network don't seek optimal flight in the classic sense. Instead, we adjust the weights (using high and low Freq. stims) in the network to produce that result.

It is a very simple system and our only interest in it is in terms of those changes within the network and the possibility to extend it to more of the network than just two or three different channels.

Or, in other words, the neural net itself isn't learning anything - they're using a peripheral effect of the network to return a very basic response, and no actual learning is occurring.

Still, it's connecting neurons to electrial wires, which is a step in the right direction, just not as big a step as it first appeared.  Another case of hype over substance.

Navy approves first ever Satanist

The British Armed Forces has officially recognised its first registered Satanist, a newspaper reports.

Naval technician Chris Cranmer, 24, has been allowed to register by the captain of HMS Cumberland, based at Devonport Naval Base in Plymouth.

The move will mean that he will now be allowed to perform Satanic rituals on board the vessel.

Former Tory minister Ann Widdecombe said she was "utterly shocked" by the Royal Navy's decision.

"Satanism is wrong. Obviously the private beliefs of individuals anywhere, including the armed forces, are their own affair but I hope it doesn't spread."

She added: "The Navy should not permit Satanist practices on board its ships.

"God himself gives free will, but I would like to think that if somebody applied to the Navy and said they were a Satanist today it would raise its eyebrows somewhat."

A spokesman for the Royal Navy said: "We are an equal opportunities employer and we don't stop anybody from having their own religious values."

I just pissed myself laughing.  Must now clean large pool of yellow liquid up.

From the BBC here.