I miss when Google was like a grep for the entire visible Internet. Now it tries to second-guess my search and direct me to a bunch of sites which all have identical information that isn't what I'm looking for.
We thought blogspam was bad, at least it was easy to ignore. It's hard to find authoritative sources for a number of topics, worryingly health advice is one of them.
i don’t agree with the google has better results thing. sometimes it does. most of the time it’s just that google has the site i want higher in the ordering than DDG. personally i’m fine scrolling down a little bit more. it’s rare i need to go to google for something that DDG doesn’t have at all in their results, but it does happen.
i do have to go to google for maps/directions/planning travel. a lot that’s annoying.
or you can press the gear button -> "Ai features: Manage" -> Search assist
Around that time Bing and DDG were actually better. Then LLM's came along and they started to take things seriously again. Maybe they think the OpenAI threat has abated enough to begin enshitification cycle 2.0.
Then it started serving synonyms, attempted to correct spelling, and so forth. Instead of serving up that there was 0 results, it attempted to be "helpful".
For a while, you could enable "verbatim" search, but even that has gotten corrupted.
Their search quality has deteriorated ever since that change. It is sad.
I've stopped using DDG now because of result quality. I now use a "meta" search backed by EXA, Tavily, and SearXNG in parallel. It can be agentically de-dupped or summarized as needed. Search as we knew it is done, largely because clicking through to evaluate result relevance before diving deeper sucks. Now we have agents that can do that portion and perform multiple searches, building on information in the last batch, to collect good results