I noticed it as well. Earlier this year I decided to self-host an aggregated search engine instead. The results page uses the various search engines, which imo improves the results (albeit not to Google's accuracy in the 2010s).
Since I don't use a proxy to hide my searches, my workflow is primarily to aggregate across privacy-friendly search engines that don't track my IP address (mwmbl, mojeek, qwant) for basic queries. For more complex queries, I use a publicly-hosted instance to aggregate across the 'better but more privacy intrusive' search engines, including Google. Because I'm using a public instance, my searches are thrown in w/ everyone else who uses the instance, which makes any effective tracking extremely unlikely (provided that you trust the instance host)
Since I don't use a proxy to hide my searches, my workflow is primarily to aggregate across privacy-friendly search engines that don't track my IP address (mwmbl, mojeek, qwant) for basic queries. For more complex queries, I use a publicly-hosted instance to aggregate across the 'better but more privacy intrusive' search engines, including Google. Because I'm using a public instance, my searches are thrown in w/ everyone else who uses the instance, which makes any effective tracking extremely unlikely (provided that you trust the instance host)
Statistics: Posted by m4c-attack — 2024-10-06 23:28