I am very wary of the “doomsday” tone of Zepnep’s piece. It reminds me of the pre-Y2K years, when a number of us were enriched by the investments companies made in software/hardware to upgrade and avoid the “doomsday” crash that was declared inevitable on January 1, 2000. We were led to believe that the stock market would crash, that flights would be grounded, that ATMs would fail, etc. (Sound familiar?) Horror still sells, I see — it sells in print and it sells consultant services too. But I think the issue is perhaps less dire. Since we can define this problem, and know well that the software we now use has a long genealogy, why can’t the development of software that addresses these problems be the next growth industry? One thing the article really tells me is that we will need human interventions for a while before software really eats the world.