Sure, I don't disagree at all that it requires a lot of engineering and good algorithms to to compete with human operators. I just think it's significantly mo

Sure, I don't disagree at all that it requires a lot of engineering and good algorithms to to compete with human operators. I just think it's significantly more of an "off the shelf" problem than the one addressed by this paper, hence my "isn't the hard part" comment. I did not mean to suggest it is something someone knocks out on a weekend.