[WikiEN-l] robotically generated content

[William Pietri]

On 04/16/2010 05:19 PM, David Gerard wrote:
> And the Cuil search engine is still the shining example of why Cuil
> Theory exists. It's comically awful and is most useful to point the
> kids at and tell them "Google got popular by not sucking like that."
>
> It is true that failure is important to future success; but one very
> important thing about failure, particularly large and spectacular
> failure is that it is - in fact - failure.
>    

Indeed, one could make the argument that Cuil's main problem is that 
they aren't hungry enough for failure.

As evidenced by things like FailCon and the Lean Startup movement, part 
of Silicon Valley culture is a taste for failing early and often as an 
approach to learning what you really should be doing. Instead of Cuil's 
approach of two giant failures, they could have been doing quiet 
releases weekly and user-testing the bejeezus out of them. Each weekly 
release would, in some sense, have been a failure, but a smaller failure 
each time.

It's no accident that Ward Cunningham, originator of the wiki, is a big 
figure in the Agile Software Development movement, which emphasizes fast 
iteration. Ward's original wiki was also the home of a lot of the early 
development of that movement. That early experience using a wiki deeply 
influenced the thinking of many of its users. Thinking that has gone on 
to infect the bulks of Silicon Valley startups. Cuil obviously aside.

William