[chc]

When somebody says "real possibility," that normally means something like "has a high enough chance of happening that it's worth considering." It is "possible" in the sense that there are things that are better than Google that could exist, but none of those things are here right now, the chances of somebody appearing out of nowhere and coming out with something so much better that everyone switches to it are practically zero. It's not a real possibility — it's just a hypothetical situation.

[snowwrestler]

This is crazy; of course it is a real possibility that someone could beat Google search. It could be a big competitor who grows over time to challenge them wholesale, or it could be a collection of smaller, smarter mobile apps that nibble away at their marketshare from every direction.

You must be fairly young...when Google originally launched, few people thought that a plain white page with a search box could unseat Altavista, let alone Microsoft.

[chc]

Wow, that is a pretty condescending reply based on an uncharitable misreading of my comment.

The point was not "Nobody will ever beat Google search no matter how much time passes." It was that Google being killed outright by a better search engine is extremely unlikely for the foreseeable future, so to call their long-term plans into question on that basis is specious.

Also, Google has not really unseated Microsoft. Microsoft's flagship products — Windows and Office — are still utterly dominant. Most of Google's battles with Microsoft have been from MS trying to encroach on Google's turf and losing. (I mean, yes, MSN Search existed before, but it was just a white-label AltaVista when Google entered the field.)

At any rate, Microsoft has been assailed on every possible front and is weakened but still massive. If you want a model of Google's future from '90s history, I'd say that's the worst case that's likely in the medium-long-term. AltaVista was a big search engine when the Web was small, but it was never Google.

[snowwrestler]

> The point was not "Nobody will ever beat Google search no matter how much time passes."

Well then I apologize--that's how I read it.

In terms of how Google might "lose", the real issue is their ad revenue growth, and I'll give an example. My employer (a nonprofit) runs online ad campaigns to raise awareness on issues. We used to spend a lot with Google; these days almost all the money goes to a combination of Facebook, Twitter, Outbrain, and Taboola. Every one of those performs more efficiently than Google.

Google could get beat like Microsoft got beat--not destroyed, but not setting the agenda either. It used to be that Microsoft could just buy or leverage their way into markets if they missed the boat early. They did it with databases, with business accounting, with the browser, with mobile (the first time around). They can't do that anymore.

Google is buying their way into the ISP market...they may not always be able to do that, even if they maintain a solid search market share.

[kgen]

Watson was demonstrated on live television, how is that not a possibility that they adapt that technology into a publicly available search engine?

[chc]

Brain-reading technology has been in use for years. It's possible that direct mental interfaces will completely obviate the keyboard, mouse and touchscreens in the near future, but it's a pretty distant possibility. As impressive as the technology is, it hasn't been shown to be anywhere near suitable for this application yet.

Who's to say that Watson's technology is even sufficiently well-suited for a search engine that it would give not only better results than Google, but results so much better that it would overcome inertia? Google has over a decade of machine learning specifically applicable to search engines. The fact that Watson does cool stuff on Jeopardy doesn't mean it's anywhere near ready to beat Google at its own game.

Simply put, these things are possible in the boolean sense of "not impossible," but are they remotely likely? No, I don't think so. They're stupendously unlikely.

[kalleboo]

IBM seem to only be interested in applying it to Medical. It's probably way too processor intensive to use as a search engine, and they can make lots of money selling hardware to hospitals instead of making pennies on ads.

[jamiek88]

Processor intensive now. In 5 years? 10?

[chc]

In 10 years, Google will almost certainly have something that takes advantage of the day's computing power as well.