I would say "overproduction", on the expectation of high valuation (i.e valuation compared to the accumulated paycheck sacrificed). I do believe there is such a thing as "too many start-ups", a problem that will self-correct.
I don't think there's a limit on the number of startups. If the economy can absorb the output of arbitrary numbers of people working in cubicles for big companies, why can't it absorb the output of arbitrary numbers of people working in little companies they've started?
FWIW, I wasn't suggesting a limit on the number of startups, so much as a practical limit on the amount of speculation that they can support.
I'd imagine that one of the things that works in YC's favor is its relative scarcity: I don't have a YC-type business in my town, let alone one with YC's community or reputation. That gives you a certain amount of leverage; you can afford to offer terms that will probably work out in your favor in the long run, and you get a nice field of applicants to cherry-pick from.
But, if the YC model starts to really take off, then at some point it's probably going to change things for you, and probably not for the better. It would be really awesome for would-be startups for a while, until investors start getting soaked and stop pouring money into it -- then we get the "Dot-com-2-point-oh" bust.
Only if they aren't profitable. Given the economic efficiency and execution speed of startups relative to freelancers/consultants and big organizations, it seems more likely they'll be the business class emerging victorious from the information revolution. It's what happened last time: those who built industrial tech that scaled beat both the cottage workers and the monarchs.
Some of the industrial giants might survive the wake of the computer, but it will probably be the same way the crowns did the engine: in name (and static, material wealth) only. Eventually the most valuable asset of the New York Times will probably be the brand and real estate.
That's an interesting thought. However, you would expect startups to be prone to more evere cycles and bubbles. People in office cubicles have more of a direct feedback from the ultimate value creation. You don't hire someone unless you need them to do a job.
Startups get started when a founder and investor are both willing to make it happen. That willingness is dependant on more speculation (in the individual case) then in the case of a cubicle worker.
Also, when you say 'startup' I expect you are talking about a specific type of starting company that fits into the YC model. While the jobs market seems to absorb an arbitrary number of people, some specific part of it such as the manufacturing sector cannot.
Not all startups are venture backed. I predict more will be side projects that turn serious. That's the logical conclusion of startups getting cheaper to build. The risk is removed, and the need to make a bet is diminished.
Sure. Success ultimately is subject to feedback. They can't succeed without creating value. But they can start without it. They may get funded without it. You may not know if you've been doing anything useful for a long time.
The cynical answer, not necessarily because I believe it fully but because nobody seems to ever bring it up:
Because the people working in companies are part of efforts that have actual an old-fashioned good or service attached. (Edit to add that this is often in a perversely indirect manner, though...)
The vast majority of startups (in the Y Combinator sense), however, are focused exclusively on connecting (be it social networking or market-making) and not on making or doing. While there's plenty of value to be contributed this way, I'd say the potential there is inherently limited.
Presumably startups produce more/better per head (that's the point, right?). On the other hand there may be a limit to what can profitably be produced by small teams. We don't nee 30 different invite or calendar apps. There are good reasons why large organizations form, and there are things that only large companies with a world- or nationwide footprint can do.
"We don't nee 30 different invite or calendar apps"
Why not? We can have two gas stations across the street from one another going after a small amount of customers. On the web we have access to 1B potential customers.
Even the "perfect" gas station -- one that had the cheapest, best tasting refreshments, the cheapest gas, the fanciest pumps, the friendliest attendants -- couldn't add new parking spaces and pumps every day in any kind of realistic manner.
But software can scale. It can scale, theoretically, infinitely, meeting the needs of every single potential user on the planet -- if the software is good enough.
