Tootfinder

Today's Metacurity kicks off with a doozy of a report on how the GRU's hackers get trained, which is followed by a string of other critical infosec developments you should know, including
--US-China weigh AI risk talks,
--Wiles says Trump won’t pick AI winners,
--Vibe-coded apps lack security,
--GothFerrari fraudster gets 78 months,
--MuddyWater masks ops as Chaos ransomware,
--Phishing hits ManageWP via Google-sponsored search ads,
--Japan u…

Russia’s hidden hacker academy exposed in massive document leak
US-China weigh AI risk talks, Wiles says Trump won’t pick AI winners, Vibe-coded apps lack security, GothFerrari fraudster gets 78 months, MuddyWater masks ops as Chaos ransomware, Phishing hits ManageWP via Google-sponsored search ads, Japan urges CI to do cyber better, much more

As for other things, like looking up medical advice, "traditional" search engines on the Internet worked much better than any LLM to get you to trustworthy sources of information.

heh, Kagi's April's Fool is actually funny: https://kagi.com

Kagi Search - A Premium Search Engine
Better search results with no ads. Welcome to Kagi (pronounced kah-gee), a paid search engine that gives power back to the user.

I am late to the party, but up to now I just thought: well, search is getting worse all over the place, and at least since Google became Alphabet it is only about the money. But “The men who killed Google”¹ explains how this was deliberate, how driven by (short-term) shareholder value increase they actively made search worse. Ant that there was an internal fight – and now Google is far away from “don't be evil”.
__
¹

The Man Who Killed Google Search
Wanna listen to this story instead? Check out this week's Better Offline podcast, "The Man That Destroyed Google Search," available on Apple Podcasts, Spotify, and anywhere else you get your podcasts. UPDATE: Prabhakar has now been deposed as head of search, read here for more details. This is the story

I am familiar with Komoot but never used it, and I guess I never will.
“Komoot has launched a ChatGPT integration…”
#noAI

Komoot Adds ChatGPT-Powered Route Discovery
Find better trails quicker & easier, with better search parameters, using this ChatGPT integration, then navigate in the komoot app.

Still, per the OP’s point, we should learn from what it is about vibe coding that really appeals to people.
The OP makes the case that we should find better abstractions and better idioms to fight boilerplate. Yes. And that we should look to things like Hypercard that reward inexperienced experimentation and exploration. Very very yes.
The latter part of my thread argues that we should •also• search for better solutions to the “Don’t make me decide! Just do something typical!” problem. I don’t know what that looks like, but we should take that problem more seriously.

Chartbeat: over two years, search traffic fell by 60% for small publishers with 1K-10K daily page views, 47% for medium-sized publishers, and 22% for large ones (Sara Fischer/Axios)
https://www.axios.com/2026/03/17/chartbeat-search-traffic-ai-chatbots

Exclusive: Small publishers hit hardest by search traffic declines
Larger publishers with better brand recognition and stronger direct-to-consumer products are more insulated from declines in traditional search traffic.

Chartbeat: over two years, search traffic fell by 60% for small publishers with 1K-10K daily page views, 47% for medium-sized publishers, and 22% for large ones (Sara Fischer/Axios)
https://www.axios.com/2026/03/17/chartbeat-search-traffic-ai-chatbots

Exclusive: Small publishers hit hardest by search traffic declines
Larger publishers with better brand recognition and stronger direct-to-consumer products are more insulated from declines in traditional search traffic.

cpe-guesser 2.0 released - Multi-Source CPE Imports, Better Ranking, and Greater Autonomy Beyond NVD
Version 2.0 brings major improvements to CPE import, ranking, and CVE v5 data handling. This release focuses on better import performance, broader format support, improved search relevance, and more robust indexing for vendor and product matching.
A notable change in this release is that cpe-guesser is no longer limited to NVD as its only practical CPE source. In addition to the N…

Nimble, whose AI agents structure real-time web data into tables that can be queried like a database, raised a $47M Series B, bringing its total funding to $75M (Ram Iyer/TechCrunch)
https://techcrunch.com/2026/02/24/nimble-way-raises-47m-to-giv…

Nimble raises $47M to give AI agents access to real-time web data | TechCrunch
Nimble uses AI agents to search the web, verify and validate the results, and then clean and structure the information into neat tables that can then be queried like a database.

@… as if you have nothing better to do … does something index-related need a nudge?
A/ seeking "FreeBSD user groups" (non-quoted) <https://mastodon.bsd…

Found an actually useful use case for "Apple Intelligence", my package tracking app uses it (fully on device) to transform Amazon product names into human-readable text; so instead of a 3-line 25-word search-optimized product name you get something like "power strip".
In this case it's also both completely inconsequential if it gets it slightly wrong and it never goes to a data center or anything.
Of course maybe it would be better if Amazon product names aren't completely out of line go start with—or even more ideally I wouldn't need to buy anything from them (I often can't for various reasons, like disability, distance to brick-and-mortar stores and prices.)

from my link log —
Arcan-A12: Weaving a different web.
https://divergent-desktop.org/blog/2020/08/10/principles-overview/
saved 2026-02-07

What really makes me irate about how LLMs are marketed and sold is that if these companies instead spend their time and money to make highly specialized versions for them we could have amazing and actually helpful tools without the ick.
This could both work much better for many use cases (for example for correlating documents and giving a list of results like a search engine instead of tedious palaver) and they wouldn't need to steal data (Professor Bender calls it succinctly "datasets too large to care")[1].
But they're pursuing "AGI" (which is provenly impossible to do with LLMs) and endless growth.
[1] https://dair-community.social/@emilymbender/116109627131276897

Investigation: Jeffrey Epstein spent hundreds of thousands on a campaign to manipulate Google search results and sanitize his Wikipedia, after his 2009 release (New York Times)
https://www.nytimes.com/2026/03/18/business/media/jeffrey-epstein-online.html

Inside Jeffrey Epstein’s Push to Cleanse His Past Online
After he left jail in 2009, Mr. Epstein hired a host of people to make him look better on Google, Wikipedia and many other places on the web.

Airbnb says its custom-built AI agent now handles ~33% of its customer support issues in North America, and it is preparing to roll out the feature globally (Sarah Perez/TechCrunch)
https://techcrunch.com/2026/02/13/airb

Airbnb says a third of its customer support is now handled by AI in the US and Canada | TechCrunch
Airbnb was poised to introduce an app that doesn't just search for you, but one that "knows you." CEO Brian Chesky said, "It will help guests plan their entire trip, help hosts better run their businesses, and help the company operate more efficiently at scale."

Immich’s search is so fucking good, it’s WAY better than Apple’s (specially after iOS 18 )

Our Call for Papers for Berlin Buzzwords closes this Sunday, February 15!
We encourage everyone in modern data infrastructure, search and machine learning and focused on open source software projects to submit their talk proposals, especially first-timers and people from underrepresented groups! #bbuzz #OpenSource #Berlin #Conference #MachineLearning #Search #DataInfrastructure #DataScience