beige.party is one of the many independent Mastodon servers you can use to participate in the fediverse.
A home to friendly weirdos. The Grey Gardens of the Fediverse (but beige). Occasionally graphically cacographic. Definitely probably not a cult (though you'll never be 100% sure). Beige-bless 🙏

Server stats:

445
active users

@TheBreadmonkey Whoa that’s a whole page of AI fakery! I didn’t imagine Google could get this bad.

@cobalt this is funny, but it's not necessarily an AI content issue. Google is just shitty at spotlighting good content now. I had a site that benefited from that.

For over a year, an event page I setup to promote a performance for the 80's band "Flock of Seagulls" kept showing up in Google's highlighted results leading us to get a puzzling amount of traffic for a topic that was barely relevant to our site.

We weren't using AI content at all. We didn't do much more than post a stock press kit bio about the band plus a publicity photo they provided. I'm sure there were dozens of almost identical pages.

I guess I should pat myself on the back for nailing the SEO so that this could happen, but there's no way a page on a regional website using boilerplate content should have been ranking anywhere near the top 5 in North America on a search like this.

@TheBreadmonkey

@sysop408 @cobalt @TheBreadmonkey yeah. So many of these “trailers” are just fan edits from the previous films. It’s pretty common, and very annoying.

People have gotten so many brainworms, they call anything they don’t understand “AI fake”

@sysop408 @cobalt @TheBreadmonkey but Google is to blame for their attempts to keep people on Google as much as possible and deprive websites of traffic by extracting all possible content from them and putting it into the search results.

When Google didn't try to be some sort of Wikipedia–Siri crossover, people had to actually click on those links and read them. So they've really brought this on themselves (and by extension, the users).

@TheBreadmonkey here's the same search in Kagi.

@jcrabapple @TheBreadmonkey Here is the same in @brave Search - no need to pay for @kagihq.

@jcrabapple @TheBreadmonkey

Kagi is becoming more and more appealing to me with this crazy AI thing around.

@henry_barreto @jcrabapple @TheBreadmonkey Even if Kagi promises privacy, I have a hard time trusting that it's 100% guaranteed when using their service requires a subscription and login.

@eighthourlunch @henry_barreto @TheBreadmonkey From their privacy policy:

Searches are anonymous and private to you. Kagi does not log and associate searches with an account.
We do not log or store your IP address. Your IP address is used only temporarily when enriching location/maps searches, and is not shared with any other party.
We only store cookies needed for site functionality.
We do not use any web browser analytics or other frontend telemetry.
We do not display any ads, or have any first-party or third-party tracking in service of ads.
We collect only the data needed to provide and protect the service.
We proxy all images to prevent tracking from third parties.
We use HTTPS encryption everywhere. All passwords are hashed and salted.

Of course, if you don't trust privacy policies or any third parties you can self-host SearXNG or Whoogle.

@jcrabapple @eighthourlunch @henry_barreto @TheBreadmonkey Ooo, that part (privacy pass) is interesting -- do they have more details on this somewhere?

(This _should_ allow users to prove to Kagi that they have a valid subscription without revealing who they are -- but of course the details are very important here ;)

@meejah @eighthourlunch @henry_barreto @TheBreadmonkey yes that is the goal, but it was just mentioned on their 2024 year end call. There should be more details coming soon.

@jcrabapple @eighthourlunch @henry_barreto @TheBreadmonkey Cool. I've previously worked on a privacy-pass based payment system, so I'm very curious how they're using it here :)

@jcrabapple @eighthourlunch @henry_barreto @TheBreadmonkey Just watched that segment: super cool!

This sounds ideal, and great use of Privacy Pass. I would still like a few more details (like when they rotate server keys, how many tokens are issued per request, and how those token-issuing requests work).

In any case, the answers to those will only slightly affect what Kagi learns; they will _not_ be able to associate searches to particular users (just that a user is legitimate).

@eighthourlunch @henry_barreto @jcrabapple @TheBreadmonkey that makes me trust them more. That means you are the customer, not advertisers.

@eighthourlunch @henry_barreto @jcrabapple @TheBreadmonkey You can use @brave Search anonymously and for free, it is independent, it doesn't rely on Google like Startpage or on Bing like DDG.

@henry_barreto @jcrabapple @TheBreadmonkey kagi is all about AI. It's funded by the same bros. And it's a US company so there is no real privacy because of us laws.

I was a paying customer until it went big on the AI crap.

The heavy, I felt, promotion of AI and remembering not to add a ? to searches (which triggers an AI summary) annoyed me.

I found this article useful
d-shoot.net/kagi.html

But sure, other companies are also bad/worse, and google is dreadful.

lori's blogWhy I Lost Faith in KagiIssues with Kagi's AI focus, finances and leadership

@swift @henry_barreto @TheBreadmonkey lol their AI is completely optional but stay angry I guess.

@henry_barreto @jcrabapple @TheBreadmonkey Kagi is all-in on LLMs. I refuse to give them money on that basis alone. plus the CEO said some pretty questionable shit in interviews recently.

@gsuberland @henry_barreto @jcrabapple @TheBreadmonkey I'm aware of the LLM thing, but can you elaborate a bit on the "some pretty questionable shit"? 🤔

@henry_barreto @jcrabapple @TheBreadmonkey

Hasn't Kagi been trying to introduce "AI" for the past year or so? They were running around talking about how it'll "humanize" searches in April, at least.

@jcrabapple @TheBreadmonkey the fediverse strikes again. Just had to explore Kagi

@jcrabapple @TheBreadmonkey How are you finding Kagi? I think I last used it at a previous company for some project, but that was about 2 years ago now.

@governorkeagan @TheBreadmonkey I love it and advocate for it. It's the way a search engine should be. It's like Google when Google was good, but it's private and more customizable.

@jcrabapple @TheBreadmonkey That’s great to hear! I’ve been using DDG for years at this point but I love the idea of customising my search engine (e.g the mastodon toggle you have)

@jcrabapple @governorkeagan @TheBreadmonkey

Kagi is about as good as Google was ten years ago. I pay for a family subscription to the basic search, have not tried their llm stuff at all.

Their maps are still developing.

@jcrabapple @TheBreadmonkey i was going to say are we never going to hear the end of people saying google sucks? it has been YEARS at this point

@jcrabapple @TheBreadmonkey wow, this is quite relevant, much better than google. I'm getting good quality results with DDG as well, but some people have been complaining that DDG is as bad for them as Google. I live in Czechie in EU so maybe some of it is regional.

@econads @TheBreadmonkey
Gad srsly what is Google doing?
They must think nobody actually needs real search anymore.
I've been training myself to add "-ai" to searches but maybe it's just time to make duckduck the default.

@grechaw @econads @TheBreadmonkey you can filter out the ai-crap from google by udm14.com/ but that doesn't hinder anyone to generate a fake trailer or, scourge of the earth, an article that uses all the right keywords but has no information at all (every time I look up if a second movie/season of something is comming)

udm14.com&udm=14 | the disenshittification Konami codeA quick way to get an AI-free search without any extra work.

@TheBreadmonkey yeah, that's my first try, too. But sometimes the information just isn't there or it's unsure if it's up to date

@grechaw @TheBreadmonkey
I've not googled anything for years, dunno what all the fuss is about ;-) not that #duckduckgo is perfect, but it's certainly useable. I have turned off the AI summaries, but that's a 1 off cookie, not something I need to do every time.

@econads @grechaw @TheBreadmonkey DDG relies on Bing results, when Bing API was recently down, DDG also didn't work, whereas Brave Search is truly independent.

@niutech @grechaw @TheBreadmonkey
Not all:
"To deliver Instant Answers on specific topics, DuckDuckGo leverages many sources, including specialized sources like Sportradar and crowd-sourced sites like Wikipedia. We also maintain our own crawler (DuckDuckBot) and many indexes to support our results. Of course, we have more traditional links and images in our search results too, which we largely source from Bing."

duckduckgo.com/duckduckgo-help

But I'm not trying to police anyone

DuckDuckGo Help PagesWhere do DuckDuckGo search results come from?DuckDuckGo is an independent internet privacy company that offers a private alternative to Google search & Chrome in one free app.

@econads @TheBreadmonkey

I've been using the DDG browser and search for a few years. Generally happy with it.

Our work laptops updated to Win11, Edge, CoPilot and Google search set as defaults. It's just an awful experience.

Fortunately, it's not locked down so much that I couldn't swap to DDG.

@Sir_Osis_of_Liver @TheBreadmonkey do companies get paid by MS to do that? What's the motivation?

@econads @TheBreadmonkey

I'm not sure.

We've got about 30k desktop/laptop Windows installs. If there was a financial benefit, I'm sure they'd go for it.

@econads @Sir_Osis_of_Liver @TheBreadmonkey It could be good deals at scale for MS licenses, and/or that IT departments select the easiest/most commonly used programs/OSes as the standard because having to troubleshoot every possibility with people who would say the browser they’re using is a “Mac” can get pretty tedious.

@WhiteCatTamer @econads @Sir_Osis_of_Liver @TheBreadmonkey it's the latter. MS doesn't give you anything for setting those defaults.

@Atemu @TheBreadmonkey
That's interesting, compared to the other kagi screenshot, it doesn't actually give you info about 5. I guess there's some personalisation behind the scenes? I know you can configure some things.

@econads @TheBreadmonkey

Yup that's it. Wikipedia is boosted for me because it's a useful factual source in most cases.

If I was into movies, I'd also likely have blocked many rumor mill shites entirely, so they wouldn't turn up at all.
You can see two sites are blocked already but I don't know or care which.

What also makes these results more bearable is that listicles ("Top 10 leaks about John Wick 5" or whatever) are filtered out entirely without needing to block the sites manually.

@Atemu @TheBreadmonkey
I looked at using kagi and while I had my 100 free searches I was using it as a backup if I couldn't find what I wanted on DDG, but I wasn't using it often enough to justify paying the monthly fee.

@TheBreadmonkey It seems day by day we are just browsing data on the internet and think we are getting information

@Nobbie @TheBreadmonkey We're not even browsing data. We're being force-fed bull crap.

@fzer0 @TheBreadmonkey I’d argue it's the internet that's broken due to generative AI, and Google is giving an accurate representation of it, but ok.

@TheBreadmonkey

Not to spoil anything, but have people even watched John Wick 4 until the end?

@airwhale @TheBreadmonkey Yeah, but Morpheus was all like "Where'd ya think he is" and Lovejoy was all like "who knows" so they definitely left it open if they wanted.

@gilester45 @airwhale

I was upset when the It monster shot General Phillip Broyles from Fringe division