Today’s (well yesterday’s, but then I got sidetracked) crazy way to save the world: place enormous arrays of windmills and wave power machines across the relatively small regions that tend to develop our major tropical storms. If the power farms are large enough, and the storm development sites can be found with enough precision, we can bleed enough energy off of developing storms to keep them from being catastrophic, while generating a lot of power off the high-activity region.
It would certainly cost a huge amount of capital, but we’d be saving money on disaster relief, and the work could be done incrementally. I think the critical question is whether we can sap off enough energy to have a significant effect. All the hurricane-fighting approaches that I’ve looked into so far attempt to stop an already fully-developed storm, which is so much a juggernaut that we don’t seem to be able to fight it. A relatively small effect, though, early enough on in the production cycle, could probably have a significant effect. And all the maps I’ve seen so far suggest that the channels in which the storms form are comparatively small, and very well known.