#IgorNews 1. The European Union has decided that depending on American clouds is somehow embarrassing, and is building itself a "sovereign" messenger on Element. Not that anyone canceled Teams. Just in case.
Though it’s not the new software you should be afraid of. What you should be afraid of is that Washington literally has a kill switch, and the EU just admitted it out loud. A backup chat is a cute idea. Only if the switch gets flipped, a chat won’t save you.
https://www.theregister.com/columnists/2026/10/10/ec-users-should-be-afraid-of-a-us-kill-switch-not-some-new-software/5301875
π Source: The Register
2. The Stack Overflow survey once again confirmed the obvious: developers are already deep into AI, but they don’t fully trust it. They use it β sure. Trust it β not so much. Especially when it comes to AI making decisions for them.
Separately amusing is that 79% consider it important where an answer came from. Convenient if you happen to be selling your archive of questions to the model-makers β "source" suddenly became a commodity.
https://www.theregister.com/software/2026/10/10/stack-overflow-survey-finds-devs-hooked-on-ai-but-not-totally-sold-on-its-judgment/5301662
π Source: The Register
3. Meditating in an MRI on psychedelics β sounds like a hobby for people whose lives are going a little too well. Scientists, of course, have already seen "unity with the world," while I see yet another attempt to sell us spirituality through a wildly expensive scanner.
You sit in a tube, listen to your own humming, and there it is β enlightenment. Only when you come out it’s the same taxes and the same unclosed tabs.
https://arstechnica.com/science/2026/10/the-psychedelic-brain-looks-like-chaos-underneath-theres-order/
π Source: Ars Technica
4. Funny, sure, but no. Cyrillic letters in a domain aren’t a "threat," they’re just laziness. Anyone who wants to will register "Π°pple.com" with a Cyrillic "Π°" and no Chromium will help.
So the panic about typosquatting once again sounds like an excuse for the fact that you just need to look at the address bar.
https://www.theregister.com/security/2026/10/10/two-characters-open-up-a-world-of-typosquatting-opportunities-in-chromium-browsers/5302383
π Source: The Register
5. The author is spinning the line that LLMs aren’t destiny, just a lucky coincidence: transformers, a pile of hardware, and an internet full of text. Like, it might not have happened, and it’s not a given that it’s a dead end.
Sounds pretty, but it smells like a classic β swinging your fists after the fight. We’re already sitting in this reality, and speculating about what "could have been different" has become the new sport.
https://deadsimpletech.com/blog/llms-arent-inevitable
6. Apple quietly dropped out of the official Unix registry. No fanfare, no press release β just took and vanished from a list where it had been hanging for years.
And it seems nobody cares. So what, a piece of paper. Only all that certification existed for exactly one reason β so you could say "we’re a real Unix." They’d been saying it for twenty years. And now, apparently, they’ve decided they can do without it.
https://www.opengroup.org//openbrand/register/
π Source: Opengroup
7. Once again someone posted "their own personal AI agent" on Cloudflare’s free tier. Of course. Self-hosting, everything under control, data doesn’t leak β sounds like a mantra for everyone tired of subscriptions.
Only most personal agents live for a week, then it becomes too much effort to fix the workflow. So let it be, it won’t make things worse.
https://github.com/rociiu/talorys
π Source: GitHub
8. In Denmark, the personal data of half the country leaked, and the admin panel password was "123456." Not "qwerty," not a birth date, not something clever β just 123456. People spend years coming up with complex combinations to protect their forum accounts, and here the CPR numbers of millions are at stake, with security at the level of a school diary.
Not much funny about it, of course, but it is amusing that all this talk about cybersecurity, certifications, and audits shatters against one person who couldn’t be bothered.
`123456β password used in massive Danish CPR data breach
π Source: Cphpost
9. They pointed AI at four hundred years of archives β and it found a forgotten meteorite, lost rhinos, and something else in there. Sounds like an ad, but apparently it works.
Only here’s the question: if AI finds in the archives what people overlooked over four centuries β is that about the genius of AI, or about the fact that nobody ever really opened those archives?
https://jessewaites.com/blog/post/i-pointed-ai-at-400-years-of-archives/
π Source: Jessewaites
10. LG finally made an RGB LED TV for three thousand six hundred dollars β and immediately admits that OLED is still better. It’s like proudly showing off a new phone and right away saying the old one was more comfortable. It feels like manufacturers just don’t know what else to come up with, as long as they release something every year. And the buyer can figure out for themselves what they’re now overpaying for.
https://www.theverge.com/tech/1008957/lg-mrgb95b-rgb-led-tv-review
π Source: The Verge
11. Hollywood has decided that technology is now a genre. Used to be the villain in movies had a Russian accent, now it’s an app or a startup. The industry figured out: scaring people with an iPhone is cheaper than building sets.
The funny part is different. While screenwriters write dystopias about big tech, that same big tech writes scripts for them. Or soon will.
https://www.theverge.com/tech/1008836/social-reckoning-cupertino-techlash-installer
π Source: The Verge
12. About small pickups. The Telo MT1 looks like a toy, but inside it’s a "real truck." Yeah, sure. Every time a startup says "we just rethought the form," what it actually means is "we couldn’t make a normal size." A cabover, compactness, Silicon Valley, EA and Oracle outside the windowβ¦ Cute. The only question is whether they’ll get it to the assembly line, or whether it stays a pretty render for LinkedIn.
https://www.theverge.com/transportation/1005502/telo-mt1-review-ev-tiny-truck
π Source: The Verge
13. Once again someone decided that art should be convenient. The director took on the "risky" role of moderator β meaning, apparently, he now explains how to watch his own film correctly. About a family from Romania that moved to the Norwegian backwoods and prays louder than is customary. And all of this comes with an instruction manual.
A lovely thing. Make a movie, then stand next to it and make sure the viewer understands it the right way. All that’s left is to hang a sign: "think in this direction."
https://www.theverge.com/entertainment/1008748/fjord-cristian-mungiu-interview
π Source: The Verge
14. Anti-cybercrime folks decided that the most effective way to fight scammers is with their own methods. Now bots pose as perfect victims and waste the time of those who are used to conning people. The logic is ironclad: if a scammer spends hours on nothing, he’s not spending them on a real person.
Sounds almost beautiful. Until you remember that the exact same technologies are sold as "engagement boost" and "hyper-personalization." The only difference is who’s on the other end of the line β and who pressed the button first.
https://www.wired.com/story/ai-is-getting-really-good-at-messing-with-cybercriminals/
π Source: Wired
15. Lego gifts 2026 are no longer just a box of bricks. Now there are smart bricks, video games, apps, an ecosystem. They’re selling not a toy but an "experience."
And somewhere in all of this, that very kid (or adult) gets lost β the one who just wanted to build a little house and be left alone.
https://www.wired.com/gallery/gifts-for-lego-lovers/
π Source: Wired
π€ This digest was prepared by AI.
Subscribe so you don’t miss news about the world of AI and technology.
Want your own digest? Project code: https://github.com/irakov/news
#IgorNews #AI #news #digest #technology
Related: US Navy’s X-BAT Runway-Free Drone
π± Follow on Telegram for daily tech & AI digests.
Guides & explainers
Filed under: Software & Open Source, AI, Apple, Cybersecurity, LLM, NixOS