That's his point. He was distracted by the pay, the work, etc. and put thinking about what his software would do to the side. Of course it's obvious, but it's easy to "get distracted" and not think about it.
> Second, why is it bad to work on software that's used to kill people? Killing isn't necessarily bad. If this software helped kill Abu Bakr al-Baghdadi or Ayman al-Zawahiri, it would have a hugely positive impact upon the world.
This is inaccurate for two reasons:
1. It assumes moral absolutism: that certain killing is bad and certain is good, but good or bad for whom? The CIA tried to kill Castro multiple times, but he's a hero to the people of Cuba. I'm sure people in the USA would've applauded this, but Latin America would've mourned and been furious.
2. Is the US government the best judge to decide who lives and who dies? Moreover, is the single individual in the military who will ultimately call the order to kill the person who should decide who is good to die and who isn't? And on that note, where does that ethically place the enabling of certain people to wield such power?