Re:Chat GPT is the birth of the real Web 3.0, and it's not g...
by skybrian on Thursday February 02, 2023 @01:35PM
(#34631098)
Some problems with that argument:
* If the variance is high enough, the average hardly matters. You're just getting slightly better noise. Example: a friend is into a sport, so does that mean you want to watch that sport? A friend watches a lot of Fox News, so does that mean you want to watch it?
* The "people I trust" graph isn't captured by "people I follow." You can follow people you don't trust, for reasons like entertainment or keeping track of what groups you dislike are up to.
* "Trust" is an abstract term that isn't well defined out of context. Trust for what purpose?
* Arguing in terms of an abstract social graph obscures whether we are talking about people or organizations and how we construct any actual graph. People follow others for whimsical reasons and often don't maintain their following list. If you ask them why they are following someone, they might not even know.
* If the variance is high enough, the average hardly matters. You're just getting slightly better noise. Example: a friend is into a sport, so does that mean you want to watch that sport? A friend watches a lot of Fox News, so does that mean you want to watch it?
* The "people I trust" graph isn't captured by "people I follow." You can follow people you don't trust, for reasons like entertainment or keeping track of what groups you dislike are up to.
* "Trust" is an abstract term that isn't well defined out of context. Trust for what purpose?
* Arguing in terms of an abstract social graph obscures whether we are talking about people or organizations and how we construct any actual graph. People follow others for whimsical reasons and often don't maintain their following list. If you ask them why they are following someone, they might not even know.