In all seriousness though, while the technology factors certainly enable some 'shrinkage' of the firm, human factors change less, which to me indicates that YC like agglomerations will not replace the corporation in many circumstances. In other words, it's easier to find people these days, but as contractors, are they really still going to slog it out for a company day in and day out for years? Most people out there are not like YC founders, would be my bet, and are reasonably happy with their 9/5 jobs, and with not having to hustle to find the next gig. On the positive side, these people will also retain the tacit knowledge inherent in many human activities, and provide a sense of continuity that a shifting network might have trouble fostering outside a small core. Apple, for instance, has 60,000+ employees, according to the web page - that's not what I'd consider a small core + network, even though they certainly outsource a lot. So... trend, yes, but fairly slow once the low-hanging fruit that tech can provide has been integrated.
The other big question I have in this vein is whether it will remain so cheap to create a company, or if it's an anomaly, and in the future more capital will be required for anything serious. My guess is that the long term is towards cheaper, but who knows what may happen.
BTW, Wikipedia says it's just an article, and here's the summary for those interested:
There's an argument for internal communications costs vs. external communications costs (Coase's thing) -- if it's more efficient to communicate within a trusted firm vs. between firms, you end up with huge firms. Things like tight-knit social networks outside of the firm (college alumni, things like YC, sites like hn, various ethnic diaspora in history, ...) would create some kind of hybrid, too, where a lot of interconnected small firms would work, so each firm could be smaller than otherwise. This would actually be interesting research for someone in economics or sociology -- I'd look at formal networks like YC and informal networks like casual coworking and web forums.)
To have some predictive value, what other industries besides web companies that are super cheap and easy to start, might morph towards this kind of structure? Or, here's another one: this industry has long been dominated by giants: IBM, Microsoft, Google, and so on. Will we see the end of that?
(And just to be clear, I think what pg has done is fantastic, and he/them are quite praiseworthy, etc... I'm just pushing back a bit on the idea. I can't foresee doing business with PG or YC, so I have no interest in not calling it like I see it)
I agree with Coase's reasoning and I'd like to agree with his prediction, but there are some empirical problems with it as of now.
Note that for any particular technology area, at first it makes sense to have a single integrated company which can redesign everything. But once the market knows roughly how to achieve the result and has standardized pieces, cooperation is much easier to achieve, and you naturally get a fragmentation into smaller, specialized, companies.
I've seen nice illustrations of the exact mechanics using examples from the computer industry. I think it was in Clayton Christensen's The Innovator's Solution, but I can't swear to that. (However that's a good book regardless.)