Croatian freediver Vitomir Maričić achieved one of the most extraordinary feats in human history by holding his breath underwater for 29 minutes and 3 seconds, a new Guinness World Record.
Experts had long considered such a duration impossible. While a trained dolphin can typically hold its breath for 8–10 minutes and most humans can barely manage one or two, Maričić remained completely still underwater for nearly half an hour on a single breath.
The attempt pushed his body to extreme limits. As the minutes passed, powerful contractions wracked his diaphragm and his organs endured intense physiological stress. Yet through years of rigorous training, mental discipline, and specialized breathing techniques, he stayed calm and focused until the end.
What makes this record even more impressive is Maričić’s purpose. He didn’t do it for fame or personal glory, he used the achievement as a powerful platform to raise global awareness about ocean conservation, marine protection, and the urgent threats facing our oceans.
A stunning demonstration of human potential and a heartfelt call to protect the planet’s most vital ecosystem.
Midjourney launched a revolutionary full-body AI medical scanner that aims to make preventive healthcare accessible transforming health monitoring to routine home health improving early disease detection, enhancing long-term health outcomes.
https://t.co/Q7MGTNckYY
BOOM!
OPEN SOURCE GLM BEATS THE FABLED FABLE!
GLM-5.2 from https://t.co/1haHftBhf1: The Open-Weight Model That Topped Claude Fable and Powers The Zero-Human Company
https://t.co/1haHftBhf1 (Zhipu AI) released GLM-5.2 and our tests show it delivering a major leap in long-horizon agentic coding with a practical 1M-token context window, flexible reasoning effort levels (High/Max), and MIT open weights.
Early benchmarks and community arenas show it excelling where it matters most for developers.
We compared it to our first Anthropic Fable model tests and GLM did better!
It leads open-weight models and has claimed the top spot on Design Arena (Elo 1360), and as I said is surpassing the now-unavailable Claude Fable 5.
It also posts strong results on coding suites: 62.1% on SWE-bench Pro (beating GPT-5.5’s 58.6) and 81.0 on Terminal-Bench 2.1.106
Official blog: https://t.co/Xt6kEOwNNK
The Zero-Human Company Goes All-In
At The Zero-Human Company, where AI agents handle nearly all operations, we’ve rolled out GLM-5.2 across all employee (agent) workflows for code generation, refactoring, debugging, and autonomous project execution. Its long-context reliability and agentic strengths make it ideal for sustained, multi-hour tasks without constant human oversight—perfect for a zero-human setup.
We’re particularly excited about its open weights and local deployment, which ensures full data privacy and resilience—no external service dependencies or potential bans.
Running GLM-5.2 Locally
Thanks to its MIT license and strong inference support, you can run GLM-5.2 (744B total params, ~40B active MoE) on your own hardware today.
Quantized versions (FP8, etc.) make it feasible on high-end setups.
Quick start options (from the official GitHub):
•vLLM: https://t.co/bkdlRlTkVS
•SGLang: https://t.co/B96vX3KnwY…/GLM-5.2
•Hugging Face Transformers or KTransformers for more options.
•Full deployment guide: https://t.co/MwHl31b7gL
Example setup with vLLM (Docker recommended for ease):
# Clone repo and follow recipes for quantized inference
# Supports reasoning_effort="max" (default) or "high"
This local-first approach aligns perfectly with our zero-human philosophy: agents run securely on-prem, with full customizability.
GLM-5.2 isn’t just competitive it’s a timely open alternative in a world of access restrictions.
We’re thrilled to test and build with it company-wide. Expect more updates as our AI workforce puts it through real production.
The myth of Mythos and the fable of Fable is entertaining but we are getting to work.
این یک سند تاریخی و پیامی از ایران مقتدر است: صلح در سایه احترام متقابل تحقق خواهد یافت.
جمهوری اسلامی ایران به صلح جهانی با حفظ عزت و استقلال، پیشرفت و همکاری منطقهای همواره متعهد و پایبند است.
Horrific attack .. Russian response to this kind of attack in past has been pretty harsh.
Hope better sense prevails and this war ends soon like the Iran war.
BIG: Ukraine launched its largest drone attack on Moscow in two years, with Russian officials saying air defenses intercepted 194 drones targeting the capital and 555 across Russia overnight.
The strikes damaged infrastructure, sparked fires, and temporarily disrupted flights at Moscow’s airports.
Among the targets was Moscow’s Kapotnya oil refinery, where videos showed heavy smoke and explosions.
Research shows a 72-hour fast can completely rebuild your immune system.
A study from the University of Southern California has revealed that fasting for 72 hours can trigger a complete regeneration of the immune system.
Researchers found that prolonged fasting causes the body to deplete its glucose and fat reserves, forcing it into a detoxifying state that eliminates damaged cells and toxins.
When normal eating resumes, stem cells are activated, leading to the production of fresh, healthy white blood cells. This immune reboot effect was observed in both mice and humans, particularly in chemotherapy patients who showed improved immune health after fasting cycles.
Lead researcher Dr. Valter Longo explains that during fasting, a gene called PKA is suppressed—this gene must be turned off for stem cells to enter regenerative mode.
As a result, the immune system essentially clears out older, weaker components and rebuilds itself anew. While more research is needed to explore the full range of benefits across organs, the study suggests that fasting could be a powerful, drug-free strategy for enhancing immunity, especially for those with weakened systems due to aging or cancer treatment.
[Valter Longo. USC Leonard Davis School]
🚨 THE INTERNET IS EATING ITSELF ALIVE. AI IS TRAINING ON AI. AND LOSING ITS MIND.
Oxford and Cambridge researchers just proved it in Nature. They call it Model Collapse.
Here is what is actually happening.
AI companies scrape the internet to train the next generation of models. But the internet is now flooded with AI-generated content. So the next generation trains on the output of the last one.
Each cycle, something gets lost. Not the average stuff. The weird stuff. The personal stuff. The things only a human would say at 2am on a forum nobody reads.
Researchers call these the tails of the distribution. The rare ideas. The unexpected takes. The things that made reading feel alive.
Those disappear first.
What survives is the average. The safe. The expected. The same six words rearranged forever.
Then the next model trains on that. And loses more.
Ross Anderson, one of the paper's authors, put it plainly. Large language models are like fire. Useful. But they pollute the environment.
The pollution is invisible. You cannot tell which sentence was written by a human. Neither can the AI about to train on it.
Model Collapse is not a technical glitch. It is a slow erasure of everything that made the internet feel like people actually lived on it.
Let me explain in detail🧵
SENIOR IRANIAN OFFICIAL TO REUTERS: UNDER DRAFT MEMORANDUM, U.S. AGREES TO RELEASE $25 BILLION OF IRAN’S FROZEN ASSETS, INCLUDING VIA DIRECT CASH TRANSFERS, COOPERATION AMONG REGIONAL COUNTRIES, AND FINANCIAL CREDIT LINES
SENIOR IRANIAN OFFICIAL TO REUTERS: UNDER DRAFT MEMORANDUM, U.S. WILL WAIVE OIL SANCTIONS ON IRAN FOR SPECIFIED PERIOD, ALLOWING IRAN TO SELL OIL AND RECEIVE REVENUES
SENIOR IRANIAN OFFICIAL: UNDER DRAFT MEMORANDUM, IRAN TO IMMEDIATELY REOPEN STRAIT OF HORMUZ TO ALL COMMERCIAL VESSELS, WHILE U.S. WILL LIFT NAVAL BLOCKADE
If a final text is ultimately agreed upon and the Trump regime doesn't implement its side of the bargain, the status of the Strait of Hormuz will immediately return to what we have witnessed over the past three and a half months.
https://t.co/fuKTdwnR5Y
@sasha_latypova@inullvoid We're all served that dish. This is now an official release taking the facts outside the realm of conspiracy, vincicating everyone that's been saying it. Did I say be "happy"? No. Just take the win. People will still deny it, but it'll be that much harder.