Claude Code is now multiplayer. Run `/delta share` in Claude Code, and your teammate joins in their own session with your full context and uncommitted work. You're both working on the same thread, backed by DeltaDB.

AI inference providers are becoming honeypots of sensitive company data because they know more about your business than you do. Yet, you're trusting their "Zero Data Retention" that retains your data for 30 days... Here is why encrypted AI should be the standard👇

Spicenet has raised a $5M Series A led by @daofive, with participation from @babylonlabs_io and @AccVentures. We’re building the brokerage network for onchain finance. 🌶️

Zcash makes freedom without surveillance possible. Now it’s easier to use than ever. Introducing @zillions_app, a new Anoma-powered app that makes @Zcash accessible for everyone. A sleek, modern, browser-based wallet with no downloads. Try it out:

Sage is now live on NEAR AI's confidential infrastucture. Intel TDX and NVIDIA confidential computing, plus cryptographic attestation, give teams a verifiable boundary for classifying and deciding on sensitive enterprise data.

Are your shielded zcash:native quantum safe today? According to @zmanian, if you have a @zodl_app wallet, your coins should be safe. Full deep dive on all things Zcash in a new @Edge_Pod below with the Zaki and @sudo_ml from @sovright_ 👇

I’m confused by the outrage from certain people around Crosslink. At this point, it’s an R&D effort. There’s no activation proposal or ZIPs, and it isn’t on the mainnet roadmap. It’s a prototype design that we’re testing and iterating on. Research and experimentation are good for Zcash. Permissionless innovation means people can explore ideas without first securing everyone’s agreement, and the Zcash community should encourage that. Good-faith debate is how Zcash improves, and we welcome scrutiny of Crosslink’s design, assumptions, and tradeoffs. We’re working on Crosslink because we think stronger protection against reorg and rollback attacks may be a necessary security upgrade for Zcash at some point. AI or other forces could eventually challenge the security assumptions behind PoW, and our thesis is that a hybrid model may offer stronger security than either pure PoW or pure PoS. Shielded Labs’ mission is to protect the long-term security of Zcash. It’s a common theme across all our projects, including Crosslink, the Network Sustainability Mechanism (NSM), and Epoch. We also demonstrated that commitment by bringing on Taylor Hornby as a security researcher to use AI tools to identify potential vulnerabilities, which uncovered the Orchard counterfeiting bug. We want to address potential risks before they become urgent, and recent events have shown the value of being forward-looking. The burden of proof is on us. If we want Crosslink included in a future network upgrade, it’s our responsibility to demonstrate consensus. But that’s still a long way off. For now, we’re focused on improving the design, incorporating user feedback, and working toward a production-ready implementation for the community and coinholders to evaluate. There’s still a lot of work to do before we’d propose Crosslink for mainnet, and we plan to run it on testnet for an extended period of time. If you disagree with Crosslink, challenge the design and help us understand what we’re missing. We all want Zcash to succeed. Let’s focus on strengthening and protecting Zcash as a whole by giving research and experimentation room to happen. If you're interested in learning more about Crosslink or participating in the incentivized testnet, feel free to DM me.

I've been running @claudeai Duet and @ChatGPT Dots side-by-side. My impressions: - Overall takeaway: Dots feels more consumer, Duet feels better for business. - Both went all in on an iMessage-style UX, and both like to send you a barrage of short messages. This quickly gets overwhelming if you're using it for a lot of things. - Duet has multiple channels that report into one main channel, like a hybrid of Slack and Grok Bot, which helps me segment work. Dots doesn't have this. - Teaching either one how you like to work takes a lot of time and manual effort, but once it's set up to your preferences it's great. - Neither is good yet at tracking and following through on soft commitments, like "I'll get back to you shortly." - Both underlying models feel great, but it's odd not being able to choose which model I'm using. - Duet's task list is great in theory but lacking in practice. It doesn't clear finished tasks unless I tell it to, and it misses ones I'd add. Still better than Dots, which has nothing. - Using Dots in the desktop app is great. Duet has no equivalent, and it's hard to do serious work from my phone. - Duet's custom instructions carry over into regular Claude, so my personal and work preferences blur together. This is clearly the tip of the iceberg and I genuinely can't wait to see where this all goes.

There’s a really pronounced divide between people who are on the AI side of things and people who are on the biology side of things. Even among people working in AI, the divide persists based on where they’re focused. The AI side, adherents of scaling laws and the bitter lesson, believe that AI getting better means AI being more capable of novel discoveries in medicine (that go beyond discovering new drugs in well understood targets or speeding up dry lab work etc). The biology people think the idea of a bitter lesson in biology is kind of ridiculous and that regardless of AI’s capabilities it’s going to be at least 5-10 years until you could hope to reliably simulate human toxicology in silico (right now AI/ML can’t even reliably simulate in vivo studies like binding affinities or functional assays). Even the most AI pilled of them, who will list all the areas that AI can reduce the time spent on various areas, don’t think the concept of AI discovering novel targets or generating new understanding of causal biology is realistic in the near term. Most, in fact, will readily characterize the assumption that even an “ASI” level model could find cures hidden in our current understanding (without expanding that understanding with significantly more biological data) as patently false. This was a big part of why we wanted to experience it ourselves by using AI for a (very) small portion of what it can do in biology. It was both mind blowing and revealed some poignant insights into how different the problem of making AI better at something like coding is to the problems of biology. Even if it unlocked the ability to generate new molecules at a stunning pace, its simulation (in silico) of binding only had 49% accuracy. And if we wanted to improve that, we’d need to feed it back the data from the lab - data that took about 30x as long to get as it took to generate the molecules. And these were tests in a test tube (turn of phrase), imagine how long it would take if we wanted to improve the model with data that required recruiting humans and monitoring their reactions. Human biology is very complex, and adding a drug - especially a novel drug and doubly so for a novel drug acting on a novel mechanism - means mapping interactions that have nth-order-effects that we don’t have the data to predict. Virtual cells have made advancements enabled by AI, but the perturbations equate to a small fraction of the possible. This is before you even get into curing disease, especially uniquely human diseases. Brain function involves cellular identities that are unique to humans, meaning animal studies are of little use. How long does it take to understand what the side effects of chronic administration of a drug in humans are? AI can speed up analyzing the data, or designing the trial or even getting it approved…but it can’t make a 6 month trial complete in 3 months. I think its important to point out that significant advancements can be made in AI that contribute to biology, while also being realistic about how much it speeds up something as complex as curing a specific cancer or diseases of the brain.

Bet you don’t know which famous investor was one of @Zcash's earliest backers 🤔 We also ask who’s been the main buyer of $ZEC lately. Both in the clip below. Full conversation in the quoted post 👇

I recently sat down with @PJVogt to chat about North Korea’s undercover U.S. workforce. I’ve been a fan of Search Engine and Reply All for years, so being a part of an episode was a real treat. If you have some spare time, give it a listen!