Introducing the Google Workspace CLI: https://t.co/8yWtbxiVPp - built for humans and agents.
Google Drive, Gmail, Calendar, and every Workspace API. 40+ agent skills included.
PSA: @TencentGlobal is aggressively scraping the Internet to build yet another AI slop chatbot, DDoSing many websites in the process.
We've found that, as of last week, their scraping bots can now solve Cloudflare challenges and behave like real users while ignoring robots.txt. In the last 24 hours alone, our website received more than 3 million successful requests from Tencent bot IP addresses, plus another 1 million that were blocked by Cloudflare challenges.
These recurring DDoS attacks from Tencent have been going on for over a year, and we have been constantly adjusting our firewall rules to filter them while trying not to impact Tencent's real users. Because that is no longer possible, we're now fully blocking Tencent IP addresses, starting with ASN 132203. We recommend other sysadmins do the same.
Other ASNs displaying similar abusive behaviour will also be fully blocked from our services.
We'd also like to thank @Cloudflare for sponsoring us with Project Alexandria as of 2025, giving our sysadmin the tools to keep RPCS3's online services running without service disruptions.
Today we're announcing ESMFold2, an open scientific engine to power prediction, design, and discovery across protein biology.
The new model delivers state of the art performance on protein interactions, especially antibodies, a critical modality for therapeutics.
We have designed and validated miniprotein binders and single chain antibodies across five therapeutic targets that are important in cancer and immunology. We are seeing very high success rates, and affinities at levels consistent with therapeutic activity.
We’re also releasing an atlas of 6.8 billion proteins, and 1.1 billion predicted structures.
ESMFold2 is built on a state of the art language model that has been trained on billions of protein sequences.
A world model of protein biology emerges through language modeling.
We’ve used the techniques of mechanistic interpretability developed to understand large language models to understand the concepts ESM uses to represent proteins.
The model’s representation space has a compositional organization of features across scales, levels of complexity, and abstraction, that reflects and mirrors the understanding of protein biology developed through a century of empirical science.
This understanding emerges without prior knowledge, just from language modeling of protein sequences.
Language models are becoming a powerful substrate to understand and program biology.
The design of protein interactions is one of the most fundamental problems in biophysics, and has critical implications for the discovery of new medicines. A simple gradient based search with the model was able to discover high-affinity protein binders.
I'm excited by the potential this has to accelerate basic science and the understanding of proteins. And especially for the new avenues it opens up for therapeutic design and medicine.
Hi Corey, you mentioned getting DMs from every direction on this. So this reply is partly to you, partly to that broader conversation and comments on both of your posts, because some of what's flying around needs context.
It unfortunately isn't that easy. Price is primarily set by the bill of materials. I can't make parts cheaper just by wanting it harder. And there's nothing extra on CORE One, every part has a job. Strip any part and you get a different printer.
What most people would not expect and many actually think it is a solution: manufacturing in China doesn't fix this either. Take an entry-level Chinese printer apart, quote the same parts in China at 100k+ volume, the parts alone cost more than the printer sells for in the West. Where is the sense in retail below parts cost? Do this long enough and price perception is permanently skewed.
Some of what props it up: 0% interest loans, multi-year tax holidays, free land, free factories, 200% R&D tax deductions. And at the wild end, the state pays consumers back ~15% of the printer's price as a rebate. They are literally running out of ways to push more money into the propped industries.
None of that is available to any Western company manufacturing in the West, or even trying to manufacture in China. And this isn't just a 3D printing problem. Same playbook took solar, drones, batteries, EVs. Now it's working on robotics. 3D printing would be done if we wouldn't be soo stubborn💪
About the “competitiveness“ I keep seeing in the replies. God I hate that word. Part of every conversation about every industry, especially in the EU. It’s shifting the problem to western industry side. Real problem is China breaking the WTO rules the next day after joining.
So an $899 CORE One isn't a price decision. And unfortunately the "what if" framing, however hopeful, reinforces exactly the view that's hurting us in too many people's eyes. Complex and heavy topic, but tried to add some context.
The totality of evidence does not support the conclusion that dietary erythritol drives cardiovascular harm.
The paper referenced claims erythritol “adversely affects brain microvascular endothelial cell function,” but it’s an in VITRO study bathing isolated cells in erythritol for 3 hours. People are using epidemiology data to frame the results of that mechanistic, in vitro study as confirming cardiovascular risk from erythritol consumption. The referenced study is interesting and possibly defensible from a mechanistic perspective, but using it to make conclusive claims about the dangers of erythritol consumption is a different story.
Erythritol-harm supporters largely ignore that humans produce erythritol endogenously.
It’s a byproduct of the pentose phosphate pathway, and that pathway is upregulated in states of hyperglycemia, oxidative stress, and pre-existing cardiovascular disease.
The landmark paper from Nature Medicine (2023), often referenced by many who are concerned about erythritol’s safety, never controlled for this. The ARIC metabolomics study from the late 80s (before erythritol entered the food supply) already showed the erythritol-CVD association. If the signal predates dietary exposure, elevated circulating erythritol is almost certainly a biomarker of metabolic dysfunction, not necessarily the cause of it.
People with diabetes and insulin resistance endogenously overproduce erythritol and might preferentially consume erythritol-sweetened products. That’s a major confounding factor, making it essentially impossible to untangle in observational data.
The strongest causal evidence we have, Mendelian randomization, doesn’t support the conclusions many are making about the safety of erythritol consumption. Most MR analyses find no causal link between genetically-predicted erythritol levels and coronary artery disease, diabetes, or chronic kidney disease.
A 2025 MR did find associations with CVD and stroke, but the effect sizes were negligible. It’s also worth emphasizing that MR analyses can’t distinguish endogenous from exogenous erythritol. Actual human clinical trials show erythritol improved endothelial function and reduced aortic stiffness in T2D patients over 4 weeks.
🔬Study shows SARS-CoV-2 causes direct damage to heart cell mitochondria - even months after recovery - helping potentially explain Long COVID heart symptoms like chest pain, palpitations & fatigue.
Let’s break it down 🧵
Najdroższy relikt zimnej wojny, która nigdy się nie skończyła, to budynek w centrum Wiednia. I nadal działa. Rosyjski kompleks "Russencity" w Wiedniu, 9-ha obiekt nad Dunajem, ma na dachach anteny satelitarne rozpoznania SIGINT. Są skierowane na Zachód. Często zmieniają pozycję. Jedna z największych anten zmieniła orientację w przeddzień Monachijskiej Konferencji Bezpieczeństwa, najważniejszego spotkania europejskich oficjeli obronnych. Wróciła na poprzednie miejsce dzień po jej zakończeniu. Ktokolwiek nią steruje, przygotował się do konferencji lepiej niż niejedna oficjalna delegacja.
Stacja SIGINT powstała w 1983 roku. Andropow zarządzał działaniami przeciwko dysydentom, kierował KGB, następnie rządził ZSRR, a przy okazji zamówił budynek. Budynek przeżył wszystko inne. Działa dokładnie tak, jak go zaprojektowano. Tak naprawdę to sieć - bo takich komórek jest więcej. Od ambasady przy cerkwi po dawną klinikę.
Rosja utrzymuje w Wiedniu około 500 pracowników dyplomatycznych, z których nawet jedna trzecia jest podejrzewana o tajną działalność wywiadowczą. Austriacka wewnętrzna agencja wywiadowcza ostrzegła, że rosyjskie stacje SIGINT "stanowią poważne zagrożenie dla bezpieczeństwa".
Rosyjska stacja SIGINT/SATCOM w Wiedniu prowadzi przechwytywanie komunikacji w paśmie C z satelitów Eutelsat 3B, Eutelsat 10B, SES-5 i Rascom QAF1, które obsługują Afrykę i łączą się z biurami ONZ, ambasadami, polami naftowymi oraz bazami wojskowymi.
To jednak nie wszystko. Olymp-K2, rosyjski satelita zdolny do operacji zbliżeniowych (RPO – Rendezvous and Proximity Operations), wykonuje nietypowe manewry, zamiast pozostawać w stałej pozycji. RPO pozwala mu zbliżać się do zachodnich satelitów, obserwować je, a nawet zakłócać ich działanie, przejmować kontrolę lub sabotować.
Dzięki bliskiej obecności, Olymp-K2 może przechwytywać niezaszyfrowane sygnały sterujące, które kontrolują pozycję satelitów, anteny oraz zarządzanie danymi. Tego rodzaju działania mogą być uznane za naruszenie granicy aktu wojennego.
Okay but this is officially the best hiring story ever. A company in Mexico rescued an orange stray cat and decided not only to keep him… but to hire him.
They named him Engineer Miauricio and gave him the title of Emotional Support Director. His responsibilities include smiling at coworkers, gently meowing, and walking around the office making everyone’s day better.
🚨 Używanie dodatków w VS Code może prowadzić do eksfiltracji danych oraz wykonania kodu (RCE).
💻 Badacze bezpieczeństwa z OX Security wykryli szereg podatności w popularnych rozszerzeniach do VS Code. Błędy występowały w Live Server, Code Runner, Markdown Preview Enhanced oraz Microsoft Live Preview. Mimo otrzymania oficjalnych zgłoszeń, twórcy dodatków nie udzielili odpowiedzi ani nie opublikowali poprawek bezpieczeństwa.⚠️ Sprawa jest o tyle niepokojąca, że rozszerzenia zostały pobrane ponad 128 mln razy i nadal są dostępne w oficjalnym sklepie.
👉 Szczegóły: https://t.co/mvqGZUr4dP
> be me
> off internet for a little
> find some free time
> (hiding in a bush)
> check phone
> "I hope nothing has occured which could cause significant geopolitical impact, economic disruption, or long term institutional damage"
> check news
I read about OpenClaw.
I still don't understand the purpose. Can someone seriously explain to me the purpose of this AI assistant? I reviewed the different "skills" it offers and it makes literally zero sense.
This can be used to automate ... messages to other people? Like on Slack, or Discord, or Telegram? Check the weather? Do stuff with Stripe? I don't understand the purpose.
Why would I need something to automatically deal with instant messaging clients? Am I missing something? Even if you don't NEED it, why would you WANT it?
I DON'T UNDERSTAND
This is really cool. I like this code, proof-of-concept, and paper A LOT.
Basically he is modifying the raw bytes of .LNK files (Windows shortcuts) to make them perform malicious actions while also operating correctly as a .LNK file. When examined from the user they will appear completely legitimate, but it's not.
This is really, really, really cool. This is a great malware technique. I can't recall the last time I read anything on .LNK files being abused in this manner. Historically they're "hijacked", not modified at the byte level.
My only criticism is he wrote this proof-of-concept in Python (not C or C++, like a gangster).
Excellent work.
@InfZakladowy To jest vibe coding. Bez wiedzy domenowej mówienie, że się przeczytało jakiś kod w nieznanej sobie technologii to tak jakby mówić, że nie znam niemieckiego ale przeczytałem raport w nim