[ExI] Reply to an article

[Anders Sandberg]

On 18/09/2012 04:32, Keith Henson wrote:
> Far as I know, the engineers of the world have not been asked how to
> solve the problems.

"If we treated climate change as an engineering problem and not a moral 
problem, it would be solved long ago". There is plenty of truth in that: 
the issue of climate change has become dominated by a particular crowd 
of people whose outlook favours particular styles of solutions (emission 
reduction) over other styles (geoengineering, adaptation, really new 
energy sources). The reasons for that are the usual sociological and 
political ones - reduction fit the green small is beautiful view, it 
looked like a governable problem from the perspective of decisionmakers, 
it did not have the uncertainties of the other approaches etc. Add 
plenty of self-interest and stir.

Engineers jumping in and suggesting that we just *solve* the problem 
don't get much traction, partially because most of the crowd doesn't 
understand them, and partially because they don't see the messy bigger 
picture of coupled economical and political systems. It is a bit like 
trying to engineer away the problems of the US election or legal system: 
sure, there might be ways that would work better, but implementing them 
are near-impossible for constitutional, legal and political reasons. 
Engineering works best when you get to start fresh on designing 
something, not when you have to retrofit a complex recalcitrant system.

Still, geoengineering is being taken somewhat seriously these days. I 
know some researchers here in Oxford, both doing simulations and 
analysing how to handle the ethics/governance problems. And they later 
problems are nontrivial: just because something can in principle be done 
doesn't mean it would be a good idea (consider a climate tug-of-war 
between Bangladesh and Scandinavia, the fact that geoengineering that 
has maximal positive impact on China's agriculture is pretty bad for 
India, and vice versa, and the fundamental problems of properly testing 
something that affects the total and only system). Everybody who knows 
the field well tell me that it better be the last resort.

I think coming up with a better way of selling efficient space solar 
power as an idea would work far better. Besides, having that 
infrastructure up there might be useful if we need to launch a solar shade.


-- 
Anders Sandberg,
Future of Humanity Institute
Oxford Martin School
Faculty of Philosophy
Oxford University