Launching a product before ‘it is perfect’ is indeed the best approach, you’ll get the valuable public feedback and comments which can be implemented sooner than later.

“Instead of reloading the page to see the next batch of results, the page just expands on the click to show the next batch” – a nice feature, but i do believe both Microsoft and Amazon’s A9 search services had this feature 12 months ago. Glad to see you’re using it too.

Good luck, you will need it (and a lot of patience).

open source, lots of cross-pollination…. everybody talking to everybody, hive mind search…. just might be the search engine of choice in the near future

enjoy

So what do you do when all eyes are on you now? Change the approach and test the software like mad? Certain things you just can’t test for, you need the product out there! I have on occasion caved and not released because I knew people would complain, but in reality, I should have just released!

I truly respect you for the approach you took on this, even when all eyes were on you. You are 100% correct, a few months/years from now, your new venture will be all over the media and magazines, and nobody will remember the amount of work it took to get you there. So to that end, Fantastic release!

Here’s something I have been telling people for a year: when I launched Wikipedia, it was empty. Not much of an encyclopedia. And in fact, my first attempt to build a freely licensed encyclopedia (Nupedia) failed completely. But one of the things Nupedia did for us as a community was give us a couple of years to talk about how to do what we wanted to do.

So, I am just following the same process here. Open source. This is how you do it. (Ever try a pre-1.0 Linux kernel? Awful!)

The difference is, of course, now everyone watches with great expectations everything I do. :)

Well, in a couple of weeks the noise will die down, and then we can just keep our noses to the ground trying to build this thing.

I posted my thoughts here:

http://feedblog.org/2008/01/06/on-wikia/

We are still in a kind of “candy era” era of the web. The success-stories of the past 15 years have mainly been about creating virtual libaries (Amazon book search, blogs, digg), galleries (flickr), phone books (myspace, facebook), media stores (napster, audiogalaxy, piratebay), trading posts (ebay, craigslist). All that stuff is pretty cool but…

The “semantic web” has more potential to change the world. Once we have widely adopted standards for “a system of meaning” we will really start seeing changes. In other words, once there is a way to compare the meaning of documents we will be able to do things like:

- have 1000+ year conversations that NEVER rehash old opinions (pretty much impossible now, since you can’t exactly review the past 1000 years of a discussion thread)
- be able to find answers for your questions… and restrict results to match a belief set (eg: ask a question on stem cells and specify that you want to exclude “christian science” in the results)
- limit duplication of effort when you post your research (even if your research is something as “trivial” as Britney Spears’ fashion sense) by searching for articles that have the same “meaning” as your own

It seems to me that Jimmy Wale’s search effort is better suited to this kind of web, since it allows for public search algorithms. I have high hopes for the future.

Maybe Wales just doesn’t care what people say. But for most people starting a hyped Web site that is guaranteed to get massive eyeballs at its first public unveiling, offering something with more meat on the bones would’ve been the way to go.

PredictAd Search Assist has a collaborative filtering feature that learns community search patterns. If you start typing in the demo attached you will start getting autocomplete results (with contextual advertisements) – these results will optimize based on the popularity of certain search terms/phrases.

If Jimmy Wales is reading – take a close look…PredictAd is FREE and its contextual advertising platform gives the publisher something that Adwords simply doesn’t! Moreover, it’s the first Ajax service offered on the web, and it’s a cool one at that…