It's true that making something people want, if you push it too far, would cause you to make something dangerously addictive. But very few startups are that successful; 99.9% of them err on the side of not making something attractive enough.
So "Make something people want" is like "Get more exercise."
Plus, if you want to try to replace this rule with one that works in that additional .1% of cases as well, it's hard to say what it should be. "Make something people should want?" Then you're on the road to fascism (if you have a lot of power) or failure (if you're just a little startup). I suspect that "Make something people want" is optimal in the same way democracy is. As Churchill said, it's the worst form of government except all the others.
As consumers, we are in charge of our own consumption decisions. Its aggravating to see people try to pass that responsibility to others.
The most successful product, service, or even startup is going to be the one that people use. Sure, instead of building a site that serves electronic "junk food," Steve could've designed a website that cuts War and Peace into "bite size chunks." But if nobody uses it, and I have a feeling nobody will, then there will be no incentive to build it.
The best way to change the web is to change your own consumption decisions. Bothered by reddit? Don't visit it. Can't resist? Block it. If enough do that, I'm sure Steve and Alexis will respond with a "healthier menu."
Making "something people want" is always good advice. I thought the interesting issue raised by this article was the particular situation where "making something people want" involved playing to the lowest common denominator. Facile content at the expense of intellectual depth. Mars Bars instead of balanced meals.
During seed-stage development, driving users by any means possible seems like great advice. (Or, at least, great advice 99.9% of the time.) Seems like it could be an increasingly dangerous strategy over time, though, at least to the extent that it encourages you to focus on lots of users/pageviews instead of finding committed users who really need your product.
Yes, but if you stop to think about it, what you've described is 95% of what doing a startup consists of. So saying yes to that = yes, we help people with their startups.
Are you confident that YC has the knowledge and/or connections to help a startup whose customer base is fundamentally different from those you've worked with in the past? What if the customer base has shown years of resistance to following the lead of technology early adopters? Would you be able to either penetrate such a market directly or find another market whose lead they would follow? Or, how much more likely would you be to succeed at this than a hacker who applies his/her problem-solving skills to the question and tries a few things?
So "Make something people want" is like "Get more exercise."
Plus, if you want to try to replace this rule with one that works in that additional .1% of cases as well, it's hard to say what it should be. "Make something people should want?" Then you're on the road to fascism (if you have a lot of power) or failure (if you're just a little startup). I suspect that "Make something people want" is optimal in the same way democracy is. As Churchill said, it's the worst form of government except all the others.