Chris Genovese, mentioned earlier in Hack the Spew, blogs at Signal + Noise. His recent post Out of Bio-Control is a delight which includes a mini biography of Charles Valentine Riley, an entomologist that did pioneering work in the biological control of agricultural pests.

Biological control - introducing organisms into environments to thwart another organism - has had many successes and some failures where the introduced organism became a pest itself. Genovese, a statistician, seems to have good clarity about the risk and reward potentials of such introductions and conducts some thought experiments about the potential hazards of current efforts to develop genetically altered organisms for use in the biological control of pests. This is a specific case of concern about an aspect of engineered replicators, biological or not, what some have called the 'grey goo' problem where a replicator goes out of control and causes trouble of unexpected magnitude.

As Genovese notes the problem already exists with non-engineered species. Conventional breeding, especially of fast reproducing bacteria and viruses, can have the same results. It isn't clear that genetic engineering introduces new threats or greater threats than conventional and mutagenic methods, and given its increased precision may even less threatening.

But the management issues remain as does the need for careful consideration of risks and rewards. Given the potential for wide effects Genovese wonders who should be involved in decision making and what standards they should use for judgement? The post is worth reading in its entirety and the blog is worth repeated visits.