Hacker News Reader

Comment by daft_pink | Hacker News Reader

I don’t quite understand why they don’t just let trusted users manually rate search results and feed that through AI to determine ranking penalties similar but not exactly like Kagi.

View on HN

https://support.google.com/websearch/answer/9281931?hl=en

They're already doing that with paid quality raters. I suppose your question is about opening this up more widely, but that's basically giving those spammers a tool to directly influence the ranking, which is going to be even worse.

I guess I don’t believe the paid quality raters are actually doing a good job.

Probably because they prefer a ranking order that promotes sites with more of their ads or some such, and/or causes users to spend more time going through results looking at more ads.

> The goals of the advertising business model do not always correspond to providing quality search to users.

- Sergey Brin and Lawrence Page, The Anatomy of a Large-Scale Hypertextual Web Search Engine

For one because it would be quite abusable. Get a user account to trusted status and then sell it. This is in addition to any scalability issues.

For the same reason Twitter, Facebook and other anti-social media companies fired there Trust and Safety teams. The parasitic elite can't trust hoomans to not turn on them and promote sites and ideas that they consider dangerous to there rule. This is the reason A.I. has risen in prominence, and is clearly biased when talking about the class war.

Are there explicit examples of bias in AI when talking about the class war?

Reply on news.ycombinator.com