Last summer Paul Graham wrote this post about trends regarding start-up companies
Among other things he stated that it is becoming easier and cheaper to start a company and more people are also inclined to take that route. There are more start-up role models and the rest of the job market is hardly inspiring. In addition there is more money around, in more experienced hands, looking for investments. This means there are more and more start-ups.
One thing we can say for sure is that there will be a lot more startups. The monolithic, hierarchical companies of the mid 20th century are being replaced by networks of smaller companies. This process is not just something happening now in Silicon Valley. It started decades ago, and it's happening as far afield as the car industry. It has a long way to run. 
Combine more start-ups with the fact that the pace of technological development keeps accelerating (see Kurzweil, e.g) and you get an environment where all incumbents should fear for their lives.
The half-life of S&P 500 and Fortune 500 companies is falling faster and faster, and that process is only just now really picking up speed:
Apple's second life that began with the iPhone in 2007 seems already on the verge of running out of time. So, within less than 10 years a complete new-comer came to dominate the smart phone market, only to turn into a me-too company again.
Google was listed in 2004 and its stock price has increased 15-fold since then, while, e.g., Microsoft's share price stayed flat between 2004-2013. Sure, MSFT's share price has had somewhat of a revival in 2014, but the company still just looks old and almost obsolete.
I'm not a fan of the Tesla share, but it is telling that a a new car company that listed only 4 years ago has increased 20-fold in market cap, and already has more than half of General Motors' market cap.
New disruptors appear more and more frequently, and quite recent disruptors end up as victims of even younger companies. This process is accelerating thanks to the law of accelereating returns, where one generation of technology make up the tools for the next generation and so on, and a more entrepreneurial climate.
The reason I am less than impressed by most of the new high flyers on the stock market is that I have seen it all before: Sketchy business models with funky company names that live off of hype and ads and each other, but don't make any profits. Most of all, they face scores of competitors and substitutes, and scores more in the near future, if you believe Graham and Kurzweil.
Take UBER taxi service as an exemple. They are priced for world domination and high margins, whereas the taxi business is perfectly suited for lots of local players, high competition and low margins. Introducing smart booking and routing systems only emphasizes those characteristics. I foresee hundreds of competing booking models and drivers being associated with several each, going with the one with the highest referral fees that particular day. UBER will never earn back its current VC valuation, unless they manage to sell it to somebody rich and desperate, such as Alibaba or TenCent, before they smash into the brick wall of reality.