Re:Google's AI answers reducing search traffic by 40%
by conductr on Wednesday November 18, 2026 @03:33AM
(#50543506)
Problem is, nobody cares enough to do that. To use that, I mean. These things were neat in the earlier Internet when it was a toy and we were seeking information. Now we have too much information.
The whole StumbleUpon idea was huge for a while too. Didn’t age well though, it was a novelty. I think this was around the time most people switched to a Feed style of information and that sucks your available attention
Re:Google's AI answers reducing search traffic by 40%
by econ on Wednesday November 18, 2026 @04:29AM
(#50543991)
If you haven't tried it you should spin up a yacy just to see how much fun it is.
The once beautiful visualisation of nodes has turned into a glowing donut of activity. Lots of people care enough to use it. You don't have to use the network, it's already great if you only crawl the website you care about.
Ideal I think is to give it some bandwidth the first few days then tune it down to the minimum.
I only ran a dozen queries but it was memorable. I didn't see a website with "under construction" in it in the decade before nor the decade after.
It seems like an unfinished approach. (Perhaps add dns data?)
Someone should do a p2p search engine and crawl those 100 to 800 thousand new domains per day. Leave them in the index for a week or a month. Make the removed ones into a download, perhaps with some [traffic] data.