With respect to the article itself, I don't think there is a clear answer. Both algorithms and humans have their advantages and disadvantages, and this will never change.

I wrote a similar article last year called “Humans vs. Computers: The Editorial Debate” – http://www.mappingtheweb.com/2007/05/14/humans-…

Has any company explored the possible combination of algorithm + human-edited?

Cheers,
Aidan

This is a great first step, and really the sites that at least go this far are ahead of the curve (though I would argue that few sites go this far).

What I am looking for is a combination of things,

The first, fundamentally (not purely) algorithmic sites that make more use of this explicit information to improve user interaction. Think Google being able to continue doing permutations of queries for you if it sees that you are having a hard time narrowing your search properly. Or being able to recommend search queries that your Gmail contacts have used successfully to find a particular piece of information.

The second is fundamentally (not purely) social sites like say Twitter, doing a better job of both collecting this explicit data and using it to make the experience better.

At the end of the day, it's a call for software adapting itself to behavior instead of software that requires adaptation.

I hope this clarifies a little. Good points.

There's no such thing as a site that is purely algorithmic or purely “crowd-based” with no logic governing it. There are examples of sites that combine the two everywhere.