5 ms·
I love how this is a problem that a huge number of people have worked on for around 20 years and have thrown stupid amounts of money at and people just jump in
by mhoad 4y ago
I love how this is a problem that a huge number of people have worked on for around 20 years and have thrown stupid amounts of money at and people just jump in and boldly proclaim that they have the answers that just popped into their mind. Peak HackerNews.
- christophilus 4y agoI don’t think these ideas just popped into his mind. They’ve been discussed for a long time on HN, and some alternative search engines have started to incorporate them. It’s worth the experiment.
- _abox 4y agoBut a company like Google would want supreme control. They're not inclined to rely on a community for anything Wikipedia shows it can work at scale.
- mhoad 4y agoGoogle have actually followed a similar model for like ten years now with like 10,000 temp workers following this 172 manual to essentially as the basis to train a lot of their ranking models. [1] https://static.googleusercontent.com/media/guidelines.raterhub.com/en//searchqualityevaluatorguidelines.pdf https://static.googleusercontent.com/media/guidelines.raterh...
- grogers 4y agoGmail's spam filter is basically community trained by people clicking spam/not spam (at least, that was what they said many years ago, it might have silently changed). What's different about search?
- permo-w 4y agospam filter isn't their key flagship aquila product
- nonameking2026 4y agoBut not my idea - Googles from Brave is something to look into, don't know entirely how it works - but it's in this direction.
- rrdharan 4y agoGoggles