What is everyone else using instead?
I use DuckDuckGo, you can configure it as your default search engine in Firefox.
I find myself hitting the same issues with ddg as with Google. The 1st page trys to sell me the thing I want info on
You can turn sponsor off in setting or use start.duckduckgo.com
I find their ad is much more clearly marked than google.
I use DDG when coding and generally find it to be a bit nicer, but from the article it sounds like they’re prone to the same issues. Anecdotally, I have noticed non-code queries do seem to be similarly crap
ETA: You can set it as default in Chrome too, under Settings -> Search Engine
I’ve turned to devdocs.io for my API documentation needs, which saves me the step of searching at all.
Searx. That it doesn’t ignore operators legitimately arouses me slightly.
If I ever can’t find something I’m looking for using a Searx instance, I switch to something like StartPage temporarily.
As great as Searx is, on very rare occasion I find myself having to switch to a different engine.
That’s funny, I’m the reverse- Searx is my final solution. I don’t like costing the lovely people hosting it any more than I have to so I start with the duck. If ddg offers nothing, searx always has it.
As an aside, are you aware that startpage was purchased some time ago? It’s owned by an ad company now.
I use DuckDuckGo Lite in Librewolf with Ublock and NoScript (amongst other addons). That combination used with ddg’s !bangs got me covered. No ads, not even image search (just go to ddg directly for that), just content.
I’ve used searx through a docker container, and it’s a strong 2nd choice, but ddg lite is just easier to set up and gives better results imho.
All said, an LLM like a local Llama and ChatGPT, can fill the gaps for very specific searches. Sadly, the modern internet takes a lot of savvy to navigate quickly and effectively.
There’s a very good reason so many are typing in “reddit” at the ends of their searches, and that’s if their not using Chat GPT to just get them to the answer faster, and using much more computational resources to do so…it’s a big cluster fuck.
