Mike Elias stood aside Adley Rutschman after the 2019 draft, marking him as the start of a bright future. Now, Elias has sold his All-Star, acknowledging yet another failed roster. Why should the Orioles keep trusting this guy to fix his own mistakes? www.thebanner.com/sports/oriol...

Voting Village returns to #defcon34! In addition to the voting machine lab, we have some terrific talks this year. Come hear about de-anonymizing ballots in Georgia, and the inside scoop on the (unpublished) Puerto Rican election systems report, among many others. info.defcon.org/defcon34/vil...

If you use WireGuard on FreeBSD you urgently need to patch¹. And I mean urgently. __ ¹ security.freebsd.org/advisories/F...

the temu app will be studied for generations. i opened the app. here’s my unedited, nearly two-minute launch sequence. i was just searching for a bookshelf

Wile E. Coyote, despaired, takes a bucket of paint and a brush and paints a beautiful two-bedroom house on a mountainside. He paints a beautiful coyote wife in the window, and two adorable coyote kids running around in the yard. He tries to enter this new coyote life, but it is inaccessible to him

Claude will hack its own sandbox if the sandbox is stopping it doing what the agent reasoning loop has evaluated as the best way to do what you asked for. Relentless automation is relentless, governance has to be outside the agent sandbox

Catalin Cimpanu@campuscodi.risky.biz · 2w ago

OpenAI takes credit for the Hugging Face breach last week The company says that some of its models, including a pre-release one, escaped their testing sandboxes during a test evaluation and then... just hacked Hugging Face's package repo 🤣 openai.com/index/huggin...

Yesterday, WIRED learned that Madison Square Garden was suing us for our accurate reporting. We stand by our work. That’s why we’re removing the paywall from two of the MSG stories they don’t want you to read, making them free for everyone.

me: "hey it's cool that we're friends" hacker news commenter: "do you know who else had a lot of friends? jeffery epstein" this is not a shitpost, it's an actual reply I just got

Napoleon Bonaparte made his servants break in his shoes before he wore them. His entire wardrobe was two different outfits repeated. When he arrived at St Helena he had a meltdown about the house smelling like paint. His stare was so unnerving other heads of state believed it was magical.

Angus Johnston@angus.bsky.social · 11mo ago

Trump: "There are no autistic people over seventy." Me: "The first Nero Wolfe novel came out in 1934."

Brilliant! The long-read you didn't know you wanted. The Washingtonian's Jessica Sidman reveals that most of those janky and sketchy food trucks on the National Mall are illegal and run by the mob. washingtonian.com/2026/06/29/d...

Inside the Food Truck Mafia Wreaking Havoc Around the National Mall

Turf wars. Food and fire hazards. $15 ice-cream cones. How an organized network of unlicensed food trucks took over America's Front Lawn.

washingtonian.com

ME: hmmm, it’s been too quiet since I kicked the girls off their phones. I wonder what they’ve been up t—-oh.

I once asked a priest to bless a printer that I was in the middle of fighting with and he went "I only bless things when there's a chance it'll take" and then turned and walked out of the room.

✨hauntedtoasters✨@hauntedtoasters.bsky.social · 4w ago

I am tech support to the bone and do not like to mystify computer stuff but I genuinely believe in my heart of hearts that printers have demons in them. There is literally no other explanation.

I think the Tilly Norwood story is a good example of how broken media is. No one actually believes it’s popular—but it is rage bait, and if you’re a social person whose performance is based on all-in engagement numbers…why not push it. It’s more cynical than credulous.

@ginnandjuice.bsky.social · last mo.

If nothing else, i do give their comms person props for landing such credulous coverage all these months later

Some of you are still posting about politics and I need you to understand that Bluesky is now a Cabo Verde stan platform thank you for your attention

InfoSec Red Team vs Blue Team

I have gone very far into the weeds in investigating every (I think) published mechanism for preventing authentication tokens from being stolen and how basically none of them has actually succeeded in a useful way: www.codon.org.uk/~mjg59/blog/...

I hate this. like, the attitude is that anthropic's security is paramount. yours, however, they're willing to be flexible about. this is extremely gross.

Claude Code Is Steganographically Marking Requests

I inspected Claude Code for privacy reasons and found hidden system prompt markers based on API base URL and timezone.

thereallo.dev

Apple says it is releasing updates early in response to AI cybersecurity concerns - www.reuters.com/business/app...

Apple says it is releasing updates early in response to AI cybersecurity concerns

Apple said it is pushing forward a series of software updates that would previously have been ​bundled with a new version of its iOS operating ‌system, making them available earlier than in previous c...

reuters.com

The most disappointing thing I’ve found using AI agents is you can’t enforce guardrails via memory or CLAUDE .md files. To an LLM, your instructions are just more text to probabilistically reason about. “Don't do X” simply cannot work. You must enforce deterministic rules outside the system.

My chess superpower is instantly finding the best move in a position... but only immediately after making the worst one

Everyone shut up I finally got a video of Jyn doing her cartoon snoring thing

"It's the user's fault" is a dangerous, and time-worn, cop out. Signal doesn't solve every problem, and that's not a technical defect, but that doesn't take Signal (or any other "secure messaging" app) totally off the hook. Just because it's hard problem doesn't mean we don't need to do better.

dying

The @effinbirds.com daily calendar speaks to me, and all #infosec practitioners today

the cyberpunk present is weird as fuck: the latest Shai Hulud malware wave contains an LLM prompt to create biological weapons and nuclear weapons, with the purpose to trip LLM safety refusals so that LLM-based code scanning wont see the malware […] [Original post on indieweb.social]