[WikiEN-l] robotically generated content

[William Pietri]

On 04/16/2010 03:09 PM, Nathan wrote:
>> >  http://www.cuil.com/info/blog/2010/04/13/cpedia-and-its-detractors
>> >
>>      
> He's not too far off the mark with some of his comments in that blog;
> it's an unfortunate side effect of the style of communication the
> Internet encourages that experimentation is often subject to vitriol
> and mocking instead of thoughtful criticism.

I think there are ways to signal that you are doing something as an 
experiment and with the requisite humility. A good example is Google 
Sets, which is part of Google Labs:

http://labs.google.com/sets

One of their earliest public experiments, they made clear that it was a 
trial with no practical purpose. Of course, the technology behind it 
surely found its way into some search-related magic. But they don't 
present it as anything more than it is. And because of that, people gave 
it 4.5 stars out of 5.

Cuil, on the other hand, took an interesting technology and pretended 
they had a Wikipedia competitor. They call it an encyclopedia, they name 
it Cpedia, and they make it look like an encyclopedia. But it's full of 
junk. So naturally, people compare it with an actual encyclopedia and 
laugh at it. And then the CEO suggests that the people criticizing them 
just don't understand the problems involved, and until they do they 
should shut up.

That is roughly the same way they failed to build a search engine, and 
it is pretty much the opposite attitude to the one that has made 
Wikipedia so strong: our incredible ambition and substantial pride is 
balanced by frankness, a deep willingness to be self-critical, and and 
ongoing invitation to others to help us improve things.

Cuil could use some of that, but I expect they'll burn through what 
remains of their $33m before they close down forever, going the way of 
Pets.com. And I'll point out that $33m is not far off from the total 
amount ever spent on Wikipedia, so it's a pretty big crater.

William