My guess is search's days are numbered and companies are "pivoting" away to other projects
a shutdown is preferable to silent bitrot
[1] Ask HN: “Tank man” image search blocked on Bing and DuckDuckGo 560 points by MaxHoppersGhost on June 4, 2021 | 126 comments | https://news.ycombinator.com/item?id=27394925
There are more or less valid arguments for not excluding Yandex[1], but as a European, I want to avoid any of my money going to Russia if possible. And there is no setting to exclude Yandex from your Kagi search results.
If you stopped using duckduckgo because of the tankman fiasco, maybe you should reconsider if Kagi is right for you.
[1] https://kagifeedback.org/d/5445-reconsider-yandex-integratio...
This is not like trying to use Bing, and then half the time you have to do the same search on Google because of how poor Bing's results are. It feels like Google felt fifteen years ago: useful results without all the "sponsored links" garbage around it.
For $10/month it’s great to have someone whose incentives are aligned with my own managing my relationships with AI companies I’d otherwise have to monitor constantly for privacy abuses.
I haven’t noticed a recent degradation in DDG’s search results, but I’m also turning to duck.ai more frequently and on the whole my search/investigation experience is better.
The one significant downside is that duck.ai limits the length of your chats, but considering the price that’s not surprising.
The direction I’d like to see the industry go is better integration of search results into AI chat, blurring the distinction between the two. That would make both products more compelling: search results are made more friendly with AI summaries, and original sources help to counter AI hallucinations and obsequious blather.
[0]: I won't bother linking any articles since there are too many articles on the subject and whatever I link is probably not the site you want (or is maybe paywalled).
Originally I switched due to their environmental focus and the way they run the company, but the quality of the results keeps me there. They have their own ! query, like DuckDuckGo. So !maps for Google Maps and !w for Wikipedia.
> We, and our search partners, collect your IP address in order to protect our service against spammers trying to conduct fraud or to up-rank specific search results.
Src.: https://www.ecosia.org/privacy
This shouldn't necessarily stop anyone; I think it should just be mentioned when it is suggested as an alternative to DuckDuckGo. You probably wouldn't switch from a search engine that proxies all favicons to avoid tracking to one that sells your identity to Google and Microsoft for tree-money.
In the last year or so, I look for the summaries from “Search Assist” and the dive into a chat with the (limited?) LLM models that it provides. It’s my go to for LLM usage. It’s rarely and for more complex needs that I go to ChatGPT.
Wonder what's different, it seems people's experience differ quite a lot.
Im expecting a future where we dont have “pages” on the internet anymore, but its just the backbone for generative AI content and if you want to promote your brand you need to pay the AI providers to put your content in responses.
Eventually, the entire notion of “searching the web” will be seem as archaic as the rotary phone.
Disclaimer: I work at Brave
To be honest, DDG has always been far behind Google. It’s fine when I know my search result is going to be in the top 10 of any engine I use, but the moment I need to search for anything non obvious I don’t even bother with DDG any more.
DDG does seem marginally worse today than it was maybe 5 years ago. It falls off rapidly past the first few results. Now it even seems like it just starts mixing generic results from some popular adjacent keyword into the results and hopes we don’t notice as users that it stopped trying to search by page 2.
And so, even if Google was the same thing it was back in 2010, there's no longer anything for "search" to find. And I hope you all downvote me to -50 and scream at me for being a retard with some snarky-assed abuse detailing how and why I am wrong. Because I don't want to be correct about this.
Right this moment it seems to work. 2 Days ago Id search for something basic like "CSS colors" and not get back a single usable result
Pretty much. Most (all?) search engines have basically stopped indexing the web. If you create content that doesn't make it through social media and has significant links, Google won't just index your website.
No, it's not under-ranking your site. It's plainly not indexing it. So if you have weird, specific content out there; it simply won't show up for a particular search.
Search is pretty much over and no one is interested in getting that fixed.
Searx is a similar idea. More powerful but definitely uglier
It is the exact opposite for me. Everyone hates the LLM based search products in my circles. Just look at this shitshow.
It doesnt require many resource and would be easy enough to run it on docker compose alongside a valkey/redis instance. I have mine on k8s but i dont think there is a helm chart easily found.
Mullvad sell a VPN and privacy-focused browser so how are they unable to proxy the searches themselves? They already have the needed tools developed.
https://developers.google.com/terms
> you will not [...] keep cached copies longer than permitted by the cache header
Privacy is an uphill battle, we should use our efforts where they make the most impact.
Still hope Russia looks like Berlin in '45 soon but until then, you can't deny they got the better search engine.
