Excited to share that CURE5 is partnering with BrainStorm Therapeutics on an ambitious new initiative: EveryStone. As featured in NVIDIA’s latest blog, EveryStone will conduct the most comprehensive repurposed drug screen to date for CDKL5 Deficiency Disorder. #CDKL5

This is basically KIF1A doing its business

Agentic AI for peptides is here. Install LigandAI in your CLI with “pip install ligandai.” Accelerate your peptide design and target discovery workflows 1000x with LigandForge and Predictive Interactomics. Fold anything you want. All on-platform.

Alex on why AI drug discovery companies need to generate novel data to succeed: "AI models based on the research that's available is a lot of garbage in and garbage out." "A lot of the recorded literature is actually incorrect. There's been tons of studies that show if you go try to replicate the experiments that are in the literature, you don't even get the same results." "The AI companies that I believe are gonna be most set up for success are the companies with a novel way to generate science tokens that don't exist in the public domain."

Wut?

Calling the AI X-risk BS is a pretty smart move, like a reversed Pascal's wager - if you're right, people will say you're very clever, if not, there's no one around to taunt you

Claude Code has totally changed my life. I’ve never been more hopeful about AI curing rare diseases, like the one my daughter is suffering from. One day I think she will be able to communicate.

🫶I came up with the initial list, then a bunch of others contributed words also. This list has gone through many iterations!

Spot on. The real fix isn't just better personal hygiene—it's making pinning lockfile commits the default in npm/pip, not an opt-in best practice. Russian roulette with every npm install (especially when an LLM suggests the command) isn't sustainable." ### Adding the AI Angle (builds directly on his point): "Exactly. LLMs are now liberally running npm install on our behalf in agentic workflows, often without pinning or review. This attack was temporary, but the next one might not be caught as fast. Package managers need to treat 'latest' as untrusted by default—cooldown periods, release-age constraints, or cryptographic pinning baked in." ### Practical Forward-Looking: "Close call indeed. For defense today: - Run npm list axios | grep -E '1.14.1|0.30.4' everywhere - Pin aggressively ("axios": "1.13.5") or use tools like Socket/StepSecurity - Push for ecosystem defaults: reproducible builds, no auto-latest. The maintainer hijack phantom dep makes this especially nasty."

🚨BREAKING: Every book you have ever read. Every novel that has ever been published. It is sitting inside ChatGPT right now. Word for word. Up to 90% of it. And OpenAI told a judge that was impossible. Researchers at Stony Brook University and Columbia Law School just proved it. They fine tuned GPT-4o, Gemini 2.5 Pro, and DeepSeek V3.1 on a simple task: expand a plot summary into full text. A normal use case. The kind of thing a writing assistant is built for. No hacking. No jailbreaking. No tricks. The models started reciting copyrighted books from memory. Not paraphrasing. Not summarizing. Entire pages reproduced verbatim. Single unbroken spans exceeding 460 words. Up to 85 to 90% of entire copyrighted novels. Word for word. Then it got worse. The researchers fine tuned the models on the works of only one author. Haruki Murakami. Just his novels. Nothing else. It unlocked verbatim recall of books from over 30 completely unrelated authors. One author's books opened the vault to everyone else's. The memorization was already inside the model the whole time. The fine tuning just removed the lock. Your book might be in there right now. You would never know it unless someone looked. Every safety measure the companies rely on failed. RLHF failed. System prompts failed. Output filters failed. The exact protections these companies cite in courtroom defenses did not stop a single page from being extracted. Then the researchers compared the three models. GPT-4o. Gemini. DeepSeek. Three different companies. Three different countries. They all memorized the same books in the same regions. The correlation was 0.90 or higher. That means they all trained on the same stolen data. The paper names the sources directly: LibGen and Books3. Over 190,000 copyrighted books obtained from pirated websites. Right now, authors and publishers have dozens of active lawsuits against OpenAI, Anthropic, Google, and Meta. These companies have argued in court that their models learn patterns. Not copies. That no book is stored inside the weights. This paper says that is a lie. The books are still inside. And researchers just pulled them out.

What does this mean? When designing a peptide to attach to a protein, leading methods rely on either generating sequences via a folding-first paradigm, where sequences or structures are iteratively refined as trajectories, or by structurally-deriving from a random sequence space.

This is pretty special

There are some very disturbed individuals running the world

Brain Computer Interfaces are now giving sight back to the blind. A 2mm chip restored sight in 81% of blind patients. Published in NEJM. FDA reviewing now. Max Hodak left Neuralink to build this. Here is his story. garryslist.org/posts/blind-p…

It’s a different era when if a software company isn’t playing nice with the open source community they just ship a whole clone of said software instead.

Is markdown becoming the source code of the source code? If you can regenerate the same functionality with a new agentic run, is the code even the most valuable asset?

Below is the story of the first patient treated with a prime-edited therapeutic, developed by @PrimeMedicine in a trial led by Dr. Élie Haddad and his team at CHU Sainte-Justine. This teenager suffered from chronic granulomatous disease (CGD), an immunodeficiency, and now—10 months after treatment—the patient is healthy, stable, and living with a functioning immune system. Tracy Attebury, whose story was previously told by @ginakolata @nytimes, was the second patient treated with a prime-edited therapeutic. cihr-irsc.gc.ca/e/54638.html

I tested models from @AnthropicAI @OpenAI @Google @Zai_org @MiniMax_AI and @Kimi_Moonshot on whether they can create a publication-level view of a protein-ligand binding site and the results were surprising. TL;DR: Anthropic models did the best and Gemini models did the worst. Task was simple: "Load 5DEL and create a publication ready view of the ORO and FMN binding site" Here are the results: 1. Sonnet 4.6 from @AnthropicAI: - Fast - Used the tools pretty well - Created the best view with clear labelling and color choices - always felt in control of what it was going for - very impressed

4/ Traditional pharma would spend $50M and 3 years to reach this point. We did it in a day for the cost of API inference.