welcome to 2026, the era of seeing "New Update!" on literally any app and creating only a sense of exasperation as you wonder what useless shit it's going to bloat you up with, or what useful shit it's going to remove I think I'll skip version thanks and check the new bug reports tomorrow x
Mer info om undersökningen: gbhackers.com/ai-hallucina...
Attackers Register AI-Hallucinated Domains to Deliver Phishing Kits and Malware
An emergent supply-chain attack vector they term "phantom squatting," in which large language models (LLMs) routinely hallucinate plausible but nonexistent domains for legitimate brands.
gbhackers.com
Anyway like I said. You don't get to label me an "AI Hater" if you can't explain how the technology works. I've closed Jira tickets older than you
Na sala ao lado, estão as métricas "curtidas" e "número de seguidores".
I mean... yes, because LLMs have no concept of "truth". They predict text. Doesn't matter what text. At no point during training are they rewarded for "truth". This is why using LLMs for information retrieval is the exact WRONG use case, but I've lost my voice screaming about this already.
Maggie Harrison Dupré@mharrisondupre.bsky.social · 2mo agoNEW: Google's AI Overviews are treating entities from the SCP Foundation — the famously fake fan-fiction universe — as if they're real. The AI feature often contextualizes SCP entities as actual discoveries while describing made-up SCP records as "official" documents. futurism.com/artificial-i...
the year is 2030 and you need a face scan, retinal scan, voice scan and to answer 20 questions about your family, health and employment history every time you use the computer. but don’t worry your personal data is safe with Butterfingers Corp in partnership with The Dark Web
I took a cruise in 2007. My first. My only (so far.) Today, I get a letter about a data breach that happened in April 2026 and offered free credit monitoring. WHY is my information still in your system to be breached? It's been 19 years.
"The purpose of a system is what it does." Something that many of my colleagues in tech conveniently forget, allowing them to escape responsibility for the real-world impact of their creations.
IMPORTANT: For any artists/writers/etc etc, using Linktree to point people to their work, from 5 July, they'll be feeding all imagery you use on your landing page into DALL-E by OpenAI. I deleted my account this morning.
Arte es Ética@arteesetica.bsky.social · 2mo ago¡ATENCIÓN! 🚨 El 5 de Julio de 2026 entra en vigencia la actualización de los ToS de Linktree. Esto incluye el uso de IA generativa (Dall-E de OpenAI) con los textos, imágenes y links que el usuario suba a su cuenta. Incluso compartirán toda la info del usuario con otros proveedores externos de IA ⚠️
AI is boring. I used to get excited about technology. Not for the last 5 years. Companies updating software that doesn't need updating simply to justify their paychecks. Adding features that are only minor changes. Turning everything into subscriptions, so they can monetize our every click.
And if anything, users are MORE unhappy about low quality software today than ever before. So if anything, we should be increasing software quality, not letting fancy autocomplete drive it into the ground.
Jason Gorman@jasongorman.bsky.social · 3mo agoA regular take I see goes something like "AI-generated code is good enough - developers need to stop being so precious about quality" Should users stop being so precious about it, too? Because the data's pretty clear that it's producing more bugs, more and longer outages, more data leaks etc
Hey! Guess what? Bitwarden is getting shitty, hope that’s not your password manager. www.fastcompany.com/91542655/bit...
Bitwarden scrubs 'Always free' and 'Inclusion' values from its website as longtime execs step down
What is going on with the beloved open-source password manager?
fastcompany.com
You REALLY should remove the tracking codes from Youtube links now. They started showing a popup about who sent the link and offering a (maybe new) DM feature, potentially revealing your personal account name to anyone who opens the link.
So if someone opts in, and has taken photos of friends, relatives, kids, even strangers, they’re “giving permission” on behalf of all those people. So lots of people have no say in this. The whole “opt in” thing was always a charade but it’s approaching terminal nonsense at this point.
Swift 6.3 Launches Official Android SDK Support Apple's Swift programming language has reached version 6.3 with a major milestone: official SDK support for Android development. The release,... #Android
Swift 6.3 Launches Official Android SDK Support
Apple’s Swift programming language has reached version 6.3 with a major milestone: official SDK support for Android development. The release, announced March 24, 2026, marks a significant expansion of Swift’s cross-platform capabilities, enabling developers to build native Android applications using a language traditionally associated with iOS. What the Android SDK Includes The Swift SDK for […]
hashlytics.io
I'm old enough to remember when software just opened and then you could start working. Now it's: wait, checking for updates, installing updates (wait I didn't want to install an update), checking for updates, please login, ask for 2FA code, login to 2FA app, verify email, why did I open this again?
But the vibe-coding guys, you don't understand. It's such a game changer.
Leadership: "AI is so good it writes almost all our code" Status dashboard:
The real problem is that the only people qualified to drive an agentic coding model are people who are intimately familiar with the codebase and *familiarity with the codebase drops when you vibe code* See, for example, the recent study commissioned by Anthropic: arxiv.org/abs/2601.20245
How AI Impacts Skill Formation
AI assistance produces significant productivity gains across professional domains, particularly for novice workers. Yet how this assistance affects the development of skills required to effectively su...
arxiv.org
