http://en.wikipedia.org/wiki/General_Motors_streetcar_conspi...
The author's point is that too many start-ups are creating fun apps that give little moments of joy rather than ones that noticeably and significantly make life easier or better for people. You can criticize the author for trying to be controversial and write something he knows has a good chance of making the front page of HN; as we've all probably noticed posts with this general sentiment are definitely in vogue right now on HN, but I think it's worth moving on from that and noting what he defines as big problems.
I think all the problems he lists we can agree are 'big' in the sense that they create a lot of confusion and inefficiency in society. In the case of medical records, 'solving' that problem could literally save countless lives.
On a related note, I disagree with the author there are big problems in society that start-ups can't solve. I think there are plenty. For many I'm guessing the idea behind 'Kill Hollywood' springs to mind. For me, it's the ubiquity of advertising in society. Personally, I'd love to see Facebook, Google, and co trying to change the world to reduce the concentration of it in our lives, unfortunately it looks like they're more interested in the opposite.
The logic of the article is that great minds are working on small problems in irrelevant startups. However, these great minds are great at certain fields, like technology. They are not particularly great at politics, for example, which is necessary for many of the problems listed in the article.
Put another way, if I want to make the most good in the world, and I'm a great programmer, what should I do, build the next Dropbox (something I'm capable of), or try and fix government regulations which cause inefficiencies in parking bills? Never mind what helps the world the most, take into account my chance of succeeding.
It's true that big problems have to be "hard" in some way, or it would be presumed that these problems would already be fixed. However, in ordinary usage, problems can be made difficult by something that is not technologically or intellectually challenging, but by something that is just practically difficult.
I think a good example would be Dwolla. Or, you might consider deposing a cruel dictator a "big problem." But it's one that was made difficult by the "artificial" challenges of rallying the right type of support, circumventing corrupt government institutions, and so on. Dealing with institutions that have made change difficult is a real and non-trivial concern, even if it's an "artificial" one.
At the same time, the Millennium Prize problems are arguably some of the "hardest problems," but people don't run around talking about how they're the world's "big problems." It isn't implausible to say that solving problems that affect people significantly may be considered "bigger problems" or at least more worth solving those that are not.
On a somewhat related note, I think some of the most important "big problems" are those which are technologically challenging and practically challenging because of current moral fashions. Many of us here may think it important to get humans to a self-sustaining colony off Earth. But most people today don't at all consider it one of the world's "big problems." Still, I think we can do it...