An interesting article. The upper continental crust deforms in a brittle manner and is amenable to modeling via Finite Element Analysis (FEA). For example McCloskey et al. (2005) used FEA to forecast an increased risk of a major earthquake on the Sumatra Subduction Zone immediately south-east of the 2004 Indian Ocean Earthquake, due to a massive build up of post-seismic stress. Sure enough, within weeks there was an 8.8 Mag earthquake in the right place.

It is possible today, and there are several other examples, where forecasts have predicted the location of increased seismic risk after a major earthquake. Large earthquakes radically change the crustal stress field, increasing the risk of earthquakes in specific areas surrounding the area of initial slip. Forecasts are useful in warning survivors and emergency workers of the location and risk of aftershocks following a major earthquake. But McCloskey et al. (2005) and Jeff McGuire are a special examples, as the structural geology of the Sumatran Subduction Zone and the East Pacific Rise are far simpler than the continental crust and were easy to model, unlike e.g. the North Anatolian fault, Turkey, where the same technique has been applied with less success.

But I agree with the article, the technique will improve when the accuracy of geological data improves. And just like weather forecasting, once we have higher resolution data earthquake forecasts will improve. We might forecast some foreshocks too one day, but that would need a dense array of instruments (I'm skeptical of electromagnetic effects preceding large earthquakes i.e. prediction). This will only ever be available in the US, Europe and Japan, it will never be a certain science and only applicable to certain fault zones. Lastly, the excessive skepticism by many in the geological community to earthquake forecasting annoys me, who are rightly skeptical of and mix it up with prediction. But this is not prediction, earthquake forecasting deserves to be accepted and researched.

McCloskey J, Nalbant SS, Steacy S (2005) Indonesian earthquake: Earthquake risk from co-seismic stress. Nature 434:291.

I have a PhD and autism, so that makes me autistic but not stupid. Simon Baron-Cohen, a autism researcher, has expressed his worry that "curing" autism could reduce the number people studying maths and other professions that require good systematizing ability, a strength possessed by people with autism. Here is his comment on the BBC website...

http://news.bbc.co.uk/1/hi/health/7736196.stm

I agree that the way to discourage curing autism, at least getting people to consider its wider implications that go beyond autism, is to connect the search into a cure with the search for genes that code for personality traits.

It is known that people with Autism and Asperger's are far more likely to vote in certain political directions and express a different degree of religiosity, so we are looking at personalty traits - of all people not just autistic people - when we look for a cure. It is scary stuff, the general public does not understand the ethics or its wider implications.

I'm not commenting on whether or not he is innocent or guilty, I'm just stating my impression that allot of his behaviors in court that led the jurors to convict him were not related to guilt but are typical behaviors seen Asperger's - such as deficits in episodic memory. You would be surprised how many court cases hinge on jurors early impressions of the indited suspect rather then tangible evidence. Having read more about the case, I agree that it is very likely that he is guilty.