Valley that thrives on hope of next big online fad

For two reasons it is probably too early to conclude that the chance to build big new Internet businesses is over.

The first is that transitions on the web offer plenty of opportunities to slip up. A decade ago, it looked like the first generation of dotcoms had staked out all the online territory worth owning, with companies such as Ebay and Amazon owning e-commerce and Yahoo the gatekeeper to online content. But Ebay failed to adapt quickly enough as its customers decided they wanted different - and more secure - ways of shopping, while Yahoo responded late both to the rise of search and of social networks. The shift to a new computing architecture greatly amplifies the risks of faulty execution. Hand-held devices that draw on services from giant central datacentres - known by the term "cloud computing" - create new, disruptive possibilities.

The second reason to expect further disruption is that the behavior of online users evolves in unexpected ways. Not so long ago, the idea that people would get hooked on broadcasting 140-character messages, or trusting an encyclopedia that can be rewritten by anyone, would have sounded outlandish. Much current experimentation on the web is focused in two areas - helping users find useful information in ways that don't involve a standard search box, and devising location-based services that tap into the potential of mobile handsets. Large areas of online content, communication and e-commerce could be at stake.