How 'next generation' do you need for that?
The rumor hit the web early this month. Citing an anonymous source, TechCrunch’s Michael Arrington reported that Microsoft was putting together some sort of uber search team at its Silicon Valley outpost in Mountain View, California. Gathering at least twenty “rock star” developers - including 23-year-old wunderkind Sanaz Ahari …
How 'next generation' do you need for that?
Is David's search a 2D representation of a 3D search?
You do realise Google powers Ask search right?
Didn't Yahoo! run with a Google search engine for years before trying their hand at their own? Would that be a huge admittance of defeat? And what's the point in handing over the whole thing to Google? Surely that would reduce revenue from 8-cents-per-click to nearly 0-cents-per-click? As Yahoo! would be walking back with their tail between their legs, Google would have them nicely placed over a barrel! (To mix my analogies...)
Hockey stick, surely... [/pedantry]
1) Ask is not "powered by Google".
2) Instead of gradually taking over Live.com, which needs perpetual explanation and long-winded references to it as "Microsoft Live Search", Microsoft needed - and still needs - standalone branding and a sensible domain name for its new search engine - especially after years of struggling along as "MSN Search". If imagination failed them, they could have bought/taken over Search.com.
I remember similar things being said about Altavista search. (Although I'm not sure they were ever *that* dominant.)
Then a certain small upstart came along providing far better search results while Altavista was messing around with trying to be some kind of web portal.
All it takes is a fast, easy to use search engine that offers something Google doesn't in the core feature of search (no trawling through pages of affiliates or blogspam, say) to woo the tech crowd and it follows from there.
Whether the giant-slaying search algorithm exists to be uncovered, that's another thing entirely.
Well done reg, classic article title/byline combination!!!
A half-decent search engine that had a simple yes/no options to filter out all the worthless price comparison sites, user forums and navel-gazing blogs would get my attention, that's for sure
That one sailed right by when I read it. Oh well, IANAC.*
* I am not a Canadian.
With so many sites gaming the searches to insure higher placement, in addition to the paid-for placements, it's hard to see any search engine getting materially better any time soon.
On the other hand, better AI interpretation of complex human-language requests would be a very nice improvement. The ability to type something like "Find information about which diesel cars have the best overall Carnot efficency" and NOT have everything that includes all the nouns and adjectives in any position (or worse, some of the nouns and adjectives) would be a major discriminator...possibly enough to seriously challenge Google. Hell, even adding a ordering requirement and a proximty descriptor ("diesel,car,Carnot,efficency within 10 words") would be a major improvement.
However, this would remove "serendipidious" click-throughs to paying sites, reducing revenue.
There was a search engine some time ago that ranked results by the amount that the site was paying for placement - and displayed the amount with the result so you could understand how it got ranked. I don't know what happened to it: it was fairly useful when looking for purchase items, as the sites paid more for a click-through the more specific the match was. I could usually find what I wanted to buy in the first 4 or 5 results, and the merchant site usually paid less than a US dollar for the placement. Not a bad deal in my book.
(To put this in perspective: a direct marketing company gets paid between US$40 and US$150 for a successful qualified sign-up solicited through phone marketing, and about half that for a snail-mail sign-up. A couple of bucks for a customer that is ACTIVELY searching for you is not a bad price.)
In fact, I'd be willing to pay a fee (less than Lexus/Nexus!) for monthly use of a search engine that would provide even the basic ordering and proximity features I note above. Maybe that's where Yahoo! or MS are missing the boat: offer a clearly superior service for a fee that offsets the lost advert revenue. Again, this is a volume issue: US$5/mo from 1,000,000 "subscribers" is US$60M each year. Make it part of an ISP fee or something like that...well, you get the picture.
I hope to live long enough to see something like this: at that point the computer will finally be doing something useful in everyday life.