APPLE RESEARCH SCIENTIST JUST SHOWED HOW 4 MAC STUDIOS RUN A TRILLION PARAMETER MODEL LOCALLY ZERO COSTS
13:18 she shows the main thing - connect 4 Mac Studios and you get 1TB of shared memory - exactly enough to run a trillion parameter model right on your desk
Apple's library - and four machines start working as one cluster
tensor parallelism: every machine holds part of every layer - all process the same token simultaneously - speed increases 3x compared to a single device
fine-tuning: one Mac Studio processes 180 tokens per second four together process 600 and not a single byte of data leaves the room
one command in the terminal - and a trillion parameter model answers from your desk and runs 24/7
data centers took years to build to run models like this - Apple did it with four Thunderbolt cables
Human trials are underway for a drug that regrows human teeth in just 4 days.
Scientists are initiating human trials for a groundbreaking drug designed to regrow teeth naturally, potentially rendering dentures and implants obsolete.
For decades, permanent tooth loss has been considered an irreversible part of aging or injury, leaving millions reliant on synthetic implants and dentures. However, researchers at Kyoto University are now challenging this biological limitation by launching human trials for a drug that triggers natural tooth regrowth. The treatment works by inhibiting a protein called USAG-1, which normally suppresses the maturation of "extra" tooth buds. By blocking this biological brake, scientists have successfully grown new, fully functional teeth in animal models, effectively tapping into a latent "third dentition" that humans carry in a suppressed state.
The project is being advanced by the startup Toregem BioPharma, with an initial clinical focus on individuals with congenital tooth agenesis—a condition where adult teeth never form. These patients are ideal candidates because their dormant tooth buds are already present but inactive, waiting for the correct biological signal to mature. While the drug requires precise timing and developmental context to work effectively, its success could signal a massive paradigm shift in restorative medicine. Moving beyond porcelain and titanium, the future of dentistry may lie in the body’s own inherent ability to regenerate its natural anatomy.
source: Takahashi, K., Kiso, H., & Murashima-Suginami, A. Advances in tooth agenesis and tooth regeneration. Regenerative Therapy
The creator of Linux just publicly called out the AI hype. Word for word.
Linus Torvalds took the stage at Open Source Summit 2026 and said this:
"When I see people saying 99% of our code is written by AI, I literally get angry. Because those same people — I can pretty much guarantee — 100% of their code is written by compilers. But they never say that."
He is not anti AI. The Linux kernel saw a 20% jump in submissions this release because of AI tools. He uses it. He gets it.
His point is something most people are too afraid to say.
AI is a productivity tool exactly like compilers were. Compilers boosted programming by 1000x. AI adds another 10x on top. Enormous. But nobody says "the compiler wrote my code." So why are we saying AI wrote it?
He also flagged something nobody is talking about.
AI is flooding small open source projects with drive-by bug reports. Someone runs a prompt, files a report and disappears when asked for a patch. Maintainers with one or two people are drowning trying to keep up.
"Sometimes AI reports a bug and when you ask for more information the person has done that drive-by and does not even answer your question. That is the real burnout issue."
And his final warning was the sharpest of all.
"People who do not understand the complexity of systems will prompt systems and write processes that will fail."
The AI hype crowd is very loud right now.
Linus has been building real systems for 35 years. When he talks, engineers listen.
Full interview here:
https://t.co/LmXJtvKc4O
BREAKING: Today, @ENERGY awarded a combined $94 million to eight companies to support deployment of advanced light-water small modular reactors (SMRs). https://t.co/Bou3jzIunY
Forget killing cancer cells. South Korea just figured out how to talk them back into being normal.
Scientists at KAIST in Daejeon have done something the world has been chasing for decades.
They found a molecular switch that flips cancer cells back into healthy cells.
No chemo. No radiation. No destroying anything.
Just… reversal.
Professor Kwang-Hyun Cho and his team caught cancer in the act. That tiny window where a normal cell is on the edge of turning malignant but hasn't fully crossed over yet. They call it the "critical transition" — the same kind of jump that happens when water hits 100°C and becomes steam.
In that split-second window, the cell is unstable. Normal and cancerous at the same time.
And that's exactly where they hit the switch.
In colon cancer trials, they targeted three master genes — MYB, HDAC2, and FOXA2 — and the cancer cells didn't die.
They went back to being healthy intestinal cells. Like nothing ever happened.
The team built a digital twin of the gene network to map every move a cell makes on its way to becoming cancerous. Then they reverse-engineered the path home.
Their paper landed in Advanced Science, published by Wiley.
It's still early. Lab trials and mice. Human treatment is years away.
But the idea of curing cancer without killing a single cell is no longer science fiction.
Source: KAIST (Korea Advanced Institute of Science and Technology), published in Advanced Science journal
🚨 NEW SOLAR CELLS JUST REACHED A SHOCKING 47% EFFICIENCY
Most solar panels capture only 15–25% of sunlight. But scientists at the National Renewable Energy Laboratory and the Fraunhofer Institute for Solar Energy Systems created advanced “six-junction” solar cells that can convert over 47% of the Sun’s energy into electricity.
These futuristic cells use ultra-concentrated sunlight and special layered materials to absorb far more energy than traditional panels. Right now they’re mainly used for satellites and high-tech systems — but this breakthrough could shape the future of clean energy.
Source: Geisz, J. F., France, R. M., Schulte, K. L., Steiner, M. A., Norman, A. G., Guthrey, H. L., Young, M. S., Song, T., & Moriarty, T. Six-junction III-V solar cell with 47.1% conversion efficiency under 143 Suns concentration.
Powell: „На Западном фронте без перемен» - заявил на своем прощальном заседании , ставка остается без изменений, со счетом 8:4 - с таким счетом, впервые за долгую историю ФРС (Fed ) -федеральной резервной системы- (центрального банка США)было принято это решение🤣
for those who dont know PromptFoo is a free tool that helps developers test their AI apps the same way you test normal software.
Instead of guessing if the AI is safe and correct, you write quick tests. The tool automatically checks things like:
- Can someone trick the AI (jailbreak)?
- Does the AI leak private information?
- Is the AI saying bad or dangerous things?
- Does the AI follow rules properly?
It’s like a safety checker for AI.
Over 300,000 developers and big companies already use it.
https://t.co/fttbWlVUDH
BREAKING: Yann LeCun has raised a HUGE $1.03bn round for his new startup, in Europe's largest seed EVER!
It values Advanced Machine Intelligence (AMI) at a WHOPPING $3.5bn!
The company, based in Paris, is aiming to develop the next generation of AI models that go beyond LLMs.
It is aiming to build world models that learn abstract representations of reality, similar to the mental models humans use to reason and guide action in the physical world.
These systems predict how situations evolve, and how actions lead to consequences, so that they can plan sequences of actions under real-world constraints.
It has just announced raising $1.03bn at a $3.5bn valuation co-led by Hiro Capital, Cathay Innovation, Greycroft, HV Capital, and Jeff Bezos Expeditions.
In February it was revealed that UK-based Ineffable Intelligence was raising a $1bn seed. Now French AMI is raising a $1.03bn seed.
Both of these companies are aiming to take AI beyond LLMs, and if they succeed, could build some of the most important technology in the world.
European Tech is COOKING and we love to see it
Свободы лишились все заместители Шойгу, пришедшие с ним в Минобороны
Вечером в четверг суд отправил под домашний арест экс-первого замминистра обороны Руслана Цаликова по обвинению в создании преступного сообщества, хищении и взятках. Он стал уже третьим лишившимся свободы заместителем Шойгу из тех, кто пришел с ним в Минобороны в начале 2010-х.
Детали. Цаликову, по данным «Коммерсанта», вменяют махинации в «Военторге», чей бывший гендиректор Владимир Павлов обвиняется в создании преступного сообщества. По версии следствия, хищения на сумму более 400 млн рублей произошли при поставках в предметов личной гигиены, закупавшихся у производителя по завышенным ценам, пишет газета. Также в деле есть эпизоды с хищением денег при снабжении военных бельем, головными уборами и формой. Общий ущерб по делу превышает 2 млрд рублей.
▪️Взятки, по версии следствия, Цаликов получал в 2019 и 2020 годах от представителей компаний, которые проводили исследования для Минобороны. Отмывались деньги, как считают в СК, через вложения в недвижимость.
▪️Цаликов — уже четвертый бывший заместитель Сергея Шойгу, который лишился свободы после его ухода. Из них трое считались людьми экс-министра, поскольку пришли в ведомство вместе с ним:
🔸Сам Цаликов работал заместителем Шойгу сначала в МЧС (2007-2012), потом в подмосковной администрации (2012).
🔸Тимур Иванов работал с Шойгу с 2012 года, когда тот был губернатором Подмосковья. Был замглавы правительства региона и перешел в Минобороны в 2013 году вслед за Шойгу. В министерстве занимался обеспечением армии, строительством и капремонтом. Приговорен в 13 годам колонии за хищение.
🔸Павел Попов работал с Шойгу в МЧС, где был его замом, а затем перешел к нему в Минобороны. Там курировал научно-исследовательскую деятельность. После ухода из ведомства Шойгу был арестован по делу о получении взятки, мошенничестве, незаконном хранением оружия, служебном подлоге и превышении должностных полномочий.
🔸Из арестованных замминистра только Дмитрий Булгаков не пришел с Шойгу в Минобороны. Он стал заместителем в 2008 году, задолго до прихода в ведомство Шойгу. Булгакова сняли с поста в 2022 году, после начала мобилизации и за полтора года до ухода Шойгу с поста. В 2024 году он был арестован по делу о мошенничестве.
▪️Вместе с Шойгу в 2012 году пришел в Минобороны также Юрий Садовенко, который занимал пост заместителя министра вплоть до отставки нынешнего секретаря Совета безопасности. До этого он работал с Шойгу в Московской области, а до этого – в МЧС, где он служил с середины 1990-х, в 2002 году стал помощником министра, а в 2007 – руководителем его аппарата. В Минобороны он организовывал повседневное планирование служебной деятельности главы ведомства. В 2025 году он умер от болезни сердца. Блог «ВЧК-ОГПУ», специализирующийся на публикации компромата и сливах от силовиков, писал о подозрительной болезни Садовенко, из-за которой ему пришлось экстренно пересадить сердце.
▪️Остальные лишившиеся поста после ухода Шойгу заместители министра обороны пришли в оборонное ведомство еще до него. Николай Панков был замминистра с 2005 года, Алексей Кузьменков служил в органах материально-технического обеспечения ВС РФ с 1992 года, Татьяна Шевцова стала советником министра обороны в 2010 году, а Александр Фомин занимался военно-техническим сотрудничеством с 1994 года.
▪️Один заместитель Шойгу в Минобороны – Андрей Булыга – в ноябре 2025 года перешел в Совбез, где занял аналогичную должность. Однако его сложно назвать человеком Шойгу – его замом он стал в 2024 году, всего за несколько месяцев до перестановок. В Минобороны он работал с 2011 года.
▪️На своих постах до сих пор находятся три замминистра обороны времен Шойгу. Юнус-Бек Евкуров, который пришел в Минобороны в 2019 году, а до этого руководил Ингушетией. Виктор Горемыкин служил в Минобороны с 2000 года. А глава Генштаба Валерий Герасимов строил армейскую карьеру со времен СССР.
Perplexity just became the first AI company to truly go head-to-head with Palantir
Using Perplexity Computer (with no local setup or single LLM limitation), it was able to build me a World Monitor with real-time data to analyze the world situation, flight situation, war situation, and many more... using skills and a combination of LLMs wherever they're best.
I'm super impressed @AravSrinivas
🧬 Universal Cancer Vaccine Edge: mRNA Trial Aims to Target All Tumors
Researchers have entered human trials with an mRNA-based cancer vaccine designed to work off the shelf, meaning no personalization for each patient’s tumor is required.
Unlike most cancer vaccines that must be tailored to a person’s individual tumor mutations, this new approach aims to trigger a broad immune response capable of going after many types of solid tumors.
In pre-clinical studies, it showed striking effects in mouse models of melanoma, brain cancers and bone cancers — especially when combined with immune-checkpoint therapies — by boosting the body’s production of type-I interferons, potent immune signals that help identify and destroy tumor cells.
Now, phase 1 (and early phase) human trials are underway. For example, the UK trial of the candidate vaccine coded mRNA-4359 is being tested in patients with advanced solid tumours, including melanoma and lung cancer.
While still early and not yet proven to work across all cancers, this development gives hope that one day we might have a “universal” cancer vaccine that doesn’t require custom-making for each patient’s unique tumor.