Originally published at: https://boingboing.net/2024/05/23/what-the-heck-is-udm14-a-new-search-engine-strips-ai-junk-from-google-results.html
…
9 Likes
It’s really frustrating that we have jump through such hoops just to do normal shit on the web… 
Also, how long before Google sues the guy who developed this?
18 Likes
Ever use a de-Googled Android phone? Here’s a de-Googled Google, or as close to one as you’re going to get on the google.com domain.
It’s such a questionably fascinating idea to offer something like this, and for power searchers like myself, it’s likely going to be an amazing tool. But Google’s decision to bury it ensures that few people will use it. The company has essentially bet that you’ll be better off with a pre-parsed guess produced by its AI engine.
It’s worth understanding the tradeoffs, though. My headline aside, a simplified view does not replace the declining quality of Google’s results, largely caused by decades of SEO optimization by website creators. The same overly optimized results are going to be there, like it or not. It is not Google circa 2001—it is a Google-circa-2001 presentation of Google circa 2024, a very different site.
But if you understand the tradeoffs, it can be a great tool. Power users will find it especially helpful when doing deep dives into things. However, is there anything you can do to minimize the pain of having to click the “Web” option buried in a menu every single time?
The answer to that question is yes. Google does not make it easy, because its URLs seem extra-loaded with cruft these days, but by adding a URL parameter to your search—in this case, “udm=14”—you can get directly to the Web results in a search.
That sounds like extra work until you realize that many browsers allow you to add custom search engines by adding the
%sentry as a stand-in for the search term you put in. I use it all the time to create shortcuts to site-specific searches I regularly use. And it works great in the case of Google.
11 Likes
Honestly, I started using kagi.com a few months back (really, really not expecting to want to use it past the free trial), and don’t want to go back to a search that is focused on getting me to buy things, look at ads and be ‘engaged’. Plus, being able to blacklist whole sites (pinterest) is nice.
I just hope it stays good - I’m happy to keep paying for it if it does.
5 Likes
I do find €10 a month for search a bit steep. Maybe that looks like a good deal when one’s using Google search, but I left that behind over 10 years ago, and I’m not looking back. I use DuckDuckGo for most of my searches, and it performs quite well.
9 Likes
I employ so damn many Firefox extensions to make tumblr mostly inhabitable. The newest twitter-esque layout is horrible, and it was heartbreaking when the “unfuck dashboard” github thingie croaked. Variations of XKit are helpful, though I was intimated AF by it at first - there are a zillion settings. I finally took a big breath and did the deep menu dive, and am so glad I did.
A tumblr extension I highly recco is Tumblr Savior. It blocks posts that contain the words, phrases, usernames you specify. It also has a whitelist feature. It naturally provides the option to view the blocked posts.
Folks on stumbleupon shared various workarounds w/each other when the enshittification began to set in.
I’ve had many Gogol fix-it extensions, both on FF and chrome. They come and go as Gogol changes its code, and breaks the external fix-its.
One that has kept working since installing it a thousand years ago is uBlacklist. It allows you to block Gogol search results from websites. You can even tailor it, once you figure out how, to block sections of a site rather than the entire site. It’s a real sanity saver.
ETA: Gogol’s advanced image search has become esp pathetic. Even when you specify ‘animated’ in the ‘type of image’ menu & specify gif as the file type, you must now put ‘animated’ in the ‘all of these words’ search box. You otherwise wind up with nothing but still jpgs and webps. I’ve figured out workarounds to access the images in the results, but the lack of ‘view image’ is beyond ridiculous. It’s evidently getty’s fault, so they can go fuck themselves.
4 Likes
-
What is “verbatim results”
-
Is there a syntax for using multiple
tbs=parameters at the same time — just throwing it in twice doesn’t seem to work
4 Likes
Yup, I agree it’s quite a lot. I guess I’ve just decided it’s worth it for now. It does feel like $5 a month would get them more than double the subscribers, but I’m sure they’ve run those numbers. Or, a cheaper plan without any ai features available.
I spent some time trying out DDG, and I really wanted to like it (I even use their tracking protection fake VPN thing in my phone), but find the results I was getting pretty mediocre (enough SEO spam sites to wade through that it was noticeable). I was usable, but I found myself annoyed, even compared to Google. That said, I should probably give it another go.
Yeah, so far, I’ve found the AI summary of Google search results to be so laughably bad that I just ignore it, but I realize a lot of people aren’t going to realize that, and it sucks that it’s just going to spread more misinformation and maybe even disinformation.
On that recent post about the USPS truck getting pulled over for racing at over 100 mph, I Googled asking the top speed of that model vehicle, and the AI summary told me it was 9 mph. I followed a few of the top links to see where it could have possibly come up with that number and I couldn’t find it, so I think it just “hallucinated” it, which is something we know these kinds of AI’s do quite a bit.
1 Like
Or you can just select “Web” instead of “All” in de search mode. That’s exactly what it does. If you search something, choose “Web” and you’ll see that it adds the &udm=14 in the url.
5 Likes
Huh. That worked. Good tip. Thanks.
ETA: And welcome to BoingBoing!
2 Likes
7 Likes
I was wondering that too. I can see the results are different but I’m not sure what the difference means.
You could try a pipe, eg. &tbs=1|2 
Or &tbs[]=1&tbs[]=2
I’ve been seeing all sorts of “puzzle solving” where people take some classic puzzle and then alter or remove the actual puzzle bit, only to have the “AI” respond by solving the original puzzle and ignoring the alterations to it. As it will, because it’s pattern matching to the bits of the much-repeated original puzzle and the many solutions to it. Which shouldn’t surprise anyone, but apparently does.
My recent favorite is the “man’s surgeon is his mother” puzzle, where it was replaced with, “man gets into accident, goes to hospital, and the surgeon who is most definitely a man, says, ‘that’s my son,’ how is this possible?” and the “AI” explains that the surgeon is actually a woman and his mother - only for the usual right-wing chuds to pop up and complain that the AI has been brainwashed by “woke programming” to make this error. And not because, you know, this is exactly how “AI” fundamentally works…
3 Likes
That’s strange. We probably have very different things we’re looking for, or different strategies for searching then. Maybe that is because I start my research for many things on specific sites, so I’ll have a good idea of my search terms when I hit the wider internet. SEO sites do turn up, but I typically can recognize those from titles and the excerpt.
Sometimes I hit a spot where I think there must be something I’m missing with DDG, but a quick !g cures that in most cases.
What is annoying is that recently DDG started showing local sites that are entirely unrelated to the search once they run out of actual search results. That’s why I don’t give them my location any more.
I have no idea what that does. I use Safari and wipr on my phone, and Brave with ublock origin on desktop, and when I last inspected the traffic, it looked good enough to me.
4 Likes
1 Like
Ah “app tracking protection”. Basically it blocks connections to known trackers from other apps on your phone, which is nice. It might not be necessary on iOS, though, if that sort of thing is against the app store policies (who knows, really). It is interesting to see the reports.
Why are you sending my data to Salesforce, banking app? Who are all those random companies, Strava and Spotify?
1 Like
Ah, I was only talking about (and checking for) browser traffic. Tracking apps is not necessarily against app store policies, but because of GDPR they‘d at least need to list those services in their privacy policy, and they need to tell apple that they use the user‘s personal data.
I wonder if that DDG-VPN would work on top of a „real“ VPN.
This topic was automatically closed after 5 days. New replies are no longer allowed.
