I have pretty mixed feelings about this. What I think the article gets right is that the false sense of profundity surrounding stuff like this is really and tru

I have pretty mixed feelings about this. What I think the article gets right is that the false sense of profundity surrounding stuff like this is really and truly ridiculous. The Yo app, I think, is/was an even better example: yes, there is definitely value in having 1-bit communication, as Marc Andreessen put it. But much like this candle, the actual fractional value achieved over existing options (text messaging, or a normal candle) is just so minimal that it absolutely does not justify the $2.5mm raise (for Yo). At the same time, the tech stack involved in producing these things, especially anything social, is so incredibly complicated that, from an "effort invested" standpoint, that hefty price tag almost seems justifiable. At the core, every digital "social" system -- that is to say, any digital system that communicates between autonomous physical-world entities instead of discrete topological points on a computer network -- every social system has a common core set of problems that have to be solved, but that are so incredibly onerous that even the "silliest" of applications end up with some serious capital. I've thought (and worked) a lot on how to solve that, this enormous disconnect between consumer value added and investor prospecting, but I think in the web world, there's simply too much momentum behind existing best practices to make any real dent. So, with a touch of irony, I think it's actually IoT, which is arguably even more difficult than your standard "social" app, that has the best chance at fixing it. My point is, there's definitely a place in the world for gimmicks like this. They're fun. They're cool. That's fine! The problem we have, and one that I hope to see solved (or solve myself) is that it's so absurdly difficult to do this stuff that these ridiculous little things get so much time and effort poured into them. Making something like this should only be marginally more difficult than writing a "flickering" integer object in Python. We've added way too much bulk to that, and I think we in tech, as an industry, would do well to add a little introspective pause into this crazy trend of "connect all the things!"