Positive. The author understands the gist of what's going on. Places like YC focus on Minimum Viable Product and doing just one thing that people want. These are great and wise strategies for both investors and entrepreneurs, but there's also another term for it that's not as flattering: bottom-feeding. Going after the little bits of value here and there, not being a hunter, existing by providing the least amount of value possible. There are huge bottom-feeders! Facebook is one of them. But it's little tidbit work. Stuff like reminding me when I like to scratch my nose, or a service for recommending dog groomers.
Negative. The problem here is that the author seems to think he knows where technology is headed, and that's always a foolish bet. His view is that we should be making complex things and solving hard problems. While I agree that hard problems need to be solved, I don't think you can decide ahead of time which problems are going to be the ones that gain traction. This is like the government or VCs "deciding" that one area of effort is going to give more results than another. Nobody knows. This is the risk of startups. It's why the smart move is to do the little stuff.
More to the point, there are two kinds of innovation (to steal from Kuhn). There is "normal" innovation, where people take existing tech and put it together in new ways to make trivially more usable things than existed before. Then there is paradigm-changing innovation, where completely new things appear.
By far, normal innovation is the norm. Probably 99.9% of all innovation is normal. Maybe more. Systems change from the ground-up, not from the top-down. Eventually a bunch of little things reach critical mass and a paradigm-changing innovation occurs. But hell, most of the time you can't force the little innovations to work, much less the big ones. I love the contrarian tone, but it's just not a realistic article.