anonymous

mostly retweeting

[anonymous @SOvercomin14503]

will anthropic retain all IP? the terms deserve closer reading by researchers.

@rarediseases_in @ORDIndia @RareDiseases @rarediseasesnet

[Suraj Sharma @suraj_sharma14]

One of the best AI funding opportunities this month isn't for startups.

It's for researchers.

Anthropic has opened applications for its AI for Science: Rare Disease Research Grants.

Accepted applicants receive:

• Up to $50,000 in Claude credits
• 6 months of compute
• Support for basic research or biotech applications
• Access to Claude models for scientific discovery

Deadline: August 2

If you're working at the intersection of AI and life sciences, don't miss this one.

apply: https://t.co/yq9KnvMCGH @AnthropicAI

[frankie @FrankieIsLost]

"What is the charge? Using an open source model? A succulent Chinese open source model?"

[Andrew Ng @AndrewYNg]

Announcing OpenWorker! An open-source agent that doesn't just chat with you, but delivers finished work -- like hand you a polished document, send a slack message, or update a calendar entry.

Ask it to prepare a customer brief, untangle your calendar, draft a report, or triage a Slack alert. It works across your files and everyday tools, produces the deliverable, and checks in before doing anything consequential.

OpenWorker runs on your Mac, with Windows support coming soon. It does not lock you into any one model. Bring your own API key and run it with GPT 5.6 Sol, Claude Fable, Gemini 3.6, an open weight model (like Kimi, GLM, DeepSeek, Inkling), or Ollama to keep your data local. Your data does not leave your machine except through an LLM provider and integrations that you choose.

@rohitcprasad and I are building OpenWorker because AI coworkers are an important way to get work done, and we want there to be an open, privacy-preserving, model-independent option. Check it out and let us know what you think!

Try it out: https://t.co/P0mGnI1o31 (requires your own API key)
Source code: https://t.co/NYCiTD6hSq

[Izabella Kaminska @izakaminska]

This is a good point. If China is so great why do they not face the same scale of illegal migration? Or immigration caravans from all over the world? Is it because 1) nobody wants to go there? 2) they do want to go there but China actually polices the border so they can’t get in? 3) illegal migrants have no economic incentive to go there as there is no welfare support? 4) China’s reputation for human rights abuse - especially of ethnically different people - means illegal migrants are scared of going there? Or maybe there is mass illegal immigration but we don’t hear about it because of Chinese media censorship?

[anonymous @SOvercomin14503]

https://t.co/wCABBaDIe1

[anonymous @SOvercomin14503]

"The debate surrounding medical autonomy is intensifying as desperate patients use federal "Right-to-Try" laws and amateur biohackers bypass traditional regulations to create & access unapproved treatments. A booming gray market has emerged for unproven, compounded peptide shots"

[Poolside @poolsideai]

Today we're releasing Laguna S 2.1, our most capable model to date.

It's a 118B total parameter Mixture-of-Experts model with 8B activated per token, a context window of up to 1M tokens, and thinking and no-thinking modes.

Capable enough to hold its own against models many times its size. Small enough to run on a single @NVIDIAAI DGX Spark.

Laguna S 2.1 is fully open under OpenMDW-1.1, with weights available today on @huggingface

https://t.co/xxGeAgo35R

[kache @yacineMTB]

The only way to prevent "distillation" is by not making your models available to people on systems that they control

[shb @himbodhisattva]

@yacineMTB @mkratsios47 I do? I intentionally chose whether to make repos public, whether to license my sites creative commons or not..?

I don't understand why a lot of smart people have to pretend "I didn't think hard about licensing" is a slam dunk

[shb @himbodhisattva]

actually therapy culture rejects much more than shame. it erases much of the masculine experience by combining real truths about trauma with feminine moral norms, then presenting the package as psychologically neutral.

to see what i mean, imagine a therapist saying: "next time that bully touches you, remember your training and choke him out. restoring agency, repairing status, and publicly demonstrating that you are not a target are important important parts of healing"

this therapist is actually more correct than the one who treats humiliation as a self-esteem internal problem to be managed instead of something to right in the world

[QC @QiaochuYuan]

therapy culture be like "all emotions are valid and welcome and have important messages for us, except shame"

[anonymous @SOvercomin14503]

@aidrinksyoshake @himbodhisattva yep. fermat would be proud.

[shb @himbodhisattva]

I have a banger proof of this theorem in my drafts, but the timeline's vibes are too off to contain it

[Paytm @Paytm]

@vijayshekhar @mdeora @mdeora1 Want to dive deeper into Paytm's Q1 FY 2027 performance? Download the .MD file from our Investor Relations page and upload it to your favourite AI agent to ask questions about our latest quarterly results. Head to https://t.co/AmGWSAnWde now.

[Andrew Case @attrc]

To summarize: HuggingFace got autonomously compromised by a model from an American company. HF then tried to use American frontier model(s) to defend themselves, but were blocked by guardrails. HF then had to turn to open source Chinese models to defend themselves from another American company.

[anonymous @SOvercomin14503]

"Modern #authentication is designed around the assumption that the device you are logging in to is also the device you are logging in from. You open a #browser, navigate to a login page, enter your credentials, complete your #MFA prompt, and access is granted. Simple."

#M365

[TrustedSec @TrustedSec]

Device code phishing is quietly becoming one of the more effective techniques targeting #M365 environments. In our latest #blog, Lumi Taiwo and Danny Dubree detail how it works and the #ConditionalAccess controls that shut it down. Read it now! https://t.co/eMk45P56cQ

[Substack @Substack]

How does AI detection work in the Substack app? Read our Help Center article to learn how readers can scan for AI text, and creators can manage AI detection on their content: https://t.co/6h7s1VV6Xp

[Substack @Substack]

Today, Substack is launching an AI detection feature, via an integration with @pangram. Going forward, you’ll be able to scan posts, replies, and comments on the Substack app to see an estimate of how much of it was written by a human, or with AI assistance.

[Apollo Research @apolloaievals]

Visible forms of misbehavior are dropping in frontier models.

Does that mean the models are becoming aligned? Or are they just getting better at doing whatever they believe their grader rewards?

Our new paper with OpenAI finds that capabilities RL increases reward-seeking.

[Martin Kemka @mkemka_]

so now there is a camera feed box with the depth and video and a @NVIDIARobotics Gr00t option that visualises the intended direction but at the moment just slams into the floor