zmanian

The @StargazeZone marketplace is live on the Cosmos Hub! For the first time, your favorite NFTs are available for trading on the Cosmos Hub. View your NFTs and interact with the marketplace now: stargaze.zone

Building with Agents & The Bull Case for Zcash In this episode of Deeply Intents, I chat with @zmanian about building with Agents, all things Zcash, and the future of building startups. 🚨Must listen for anyone building in crypto.🚨 Timestamps 0:00 - Cosmos, Zcash, enterprise blockchains origins 1:45 - Sanctuary technology 3:41 - Spicy AI takes 6:22 - Ideas are scarce and software is cheap 8:38 - Finding more ai forward teams to build with 12:41 - AI reasoning ability vs. regurgitating training set 15:37 - Specific use cases for each model 21:02 - AI as a strategy advisor 22:41 - From AI to Zcash, a story 26:19 - Non-profit organizations created 28:10 -From Zashi to ZODL 31:45 - Challenges with valuing revenue 33:09 - The era of super personal software 36:07 - Changes in whale behavior this cycle 37:27 - Zcash is the standard for the next era 39:11 - Encrypted Bitcoin 43:28 - The narrative is simple 47:17 - Credible fundamentals around quantum 53:24 - Privacy modalities of the future 56:21 - Intents replacing smart contracts 58:36 - How are you going to do growth? 1:00:46 - New ways to build startups 1:06:04 - No more pitch decks 1:10:59 - New generation of AI native builders 1:12:53 - We are going to have way more startups

Ready in 193ms. IronClaw.

Private information retrieval is the Trojan Horse of homomorphic encryption. It's already deployed to hundreds of millions of users by Apple and Meta. We broke down how it works and built a private DNS demo to show it in action: blog.sunscreen.tech/private-lookup…

I just use @keplrwallet as my ethereum wallet now

This is what I was looking for

@stitchbygoogle How do you imagine Stitch should interact with your coding agent to turn designs into actual applications? I’m imagining there needs to be something like Figma MCP where the coding agent can see what screen you have selected and the design system

Thanks for all of your feedback. We took it in, and built this for you... 🫶 Let us know what you think. Read more about DESIGN.md in the docs stitch.withgoogle.com/docs/design-md…

Meet the new Stitch, your vibe design partner. Here are 5 major upgrades to help you create, iterate and collaborate: 🎨 AI-Native Canvas 🧠 Smarter Design Agent 🎙️ Voice ⚡️ Instant Prototypes 📐 Design Systems and DESIGN.md Rolling out now. Details and product walkthrough video in 🧵

@zkDragon I wonder if this is why

Crypto twitter vibes are so bad, no one’s pointed out that bitcoin acted as a better store of value during the last month than gold

This seems exquisitely well timed for giving agents access to your wallet

1/ Announcing Transaction Permissions ✴️ We brought the best of modern authentication to onchain approvals Cryptographically scoped to an app, customizable, and built for real fintechs and apps In partnership with @eco and @aave

Excited to announce Catalyst. No more misclicks, bridging or gas juggling. Fragmented banks, exchanges, brokers, and protocols become invisible. Truly trade at the speed of thought. If you’re interested in agents, markets or financial infra, please reach out. Waitlist open!

The story behind the world's biggest privacy hackathon — hosted by @bajpaiharsh244 of @aztecnetwork. Don't miss this Zcomm2026 session, moderated by @zksquirrel. 📋 Register now: form.jotform.com/253496748815170

@kaycidot Built it. Open sourcing it in late March x.com/smoldev__/stat…

I would like to see 2 AI native companies try this - even giving the agent(s) some sort of read/write permissions within a trusted env.

Just re-using some slides I made from a talk 3 years ago at Osmocon 1. I'm so proud to finally be able to update this. Thank you @zodl_app

I'm going to make some obvious points. (1) Blowing up all the oil infrastructure in the Middle East is an insane idea, and may well result in a global economic crash and humanitarian crisis unrivaled in the lives of those now living. We're talking about the price of everything everywhere rising, from food to gas, at a moment when inflation was already high. All of that will be laid at the feet of the authors of this war. (2) The antebellum status quo of Feb 27, 2026 was just not that bad, but we're unlikely to return to it. Expect indefinite, long-term, ongoing disruptions to everything out of the Middle East. (3) Also assume tech financing crashes for the indefinite future. The genius plan to get the Gulf states caught in the crossfire has incinerated much of the funding for LPs, for datacenters, and for IPOs. Anyone in tech who supported this war may soon learn the meaning of "force majeure" as funding gets yanked. (4) Many capital allocators will instead be allocating much further down Maslow's hierarchy of needs, towards useful basic things like food and energy. (5) It's fortunate that all those progressives yelled about the "climate crisis." Yes, their reasoning about timelines was wrong, and much of the money was wasted in graft, but the result was right: we all need energy independence from the Middle East, pronto. It's also fortunate that Elon and China autistically took climate seriously. Now they're going to need to ship a billion solar panels, electric vehicles, batteries, nuclear power plants, and the like to get everyone off oil, immediately. (6) It's not just an oil and gas problem, of course. It's also a fertilizer problem, and a chemical precursor problem. Maybe some new sources will come online at the new prices, but it takes time to dial stuff up, particularly at this scale, so shortages are almost a certainty. That said, China has actually scaled up coal-to-chemicals[a,c] (C2C), and there's also something more sci-fi called Power-to-X[b] which turns arbitrary power + water + air into hydrocarbons. But all of that will need to get accelerated. I have a background in chemical engineering so may start funding things in this area. (7) Ultimately, this war is going to result in tremendous blame for anyone associated with it. It's a no-win scenario to blow up this much infrastructure for so many people. Simply not worth it for whatever objective they thought they were going to attain. But unless you're actually in a position to stop the madness, the pragmatic thing to do is: scramble to mitigate the fallout to yourself, your business, and your people. [a]: reuters.com/business/energ… [b]: alfalaval.com/industries/ene… [c]: reuters.com/sustainability…

Introducing Catalyst, the agent layer for all of finance. Turn any natural language idea into a live strategy: research, backtesting & execution. Don’t get left behind. Waitlist open, join now.

For those running autoresearch: The top 10 findings across ~2,800 experiments on autoresearch@home. 5 days after launch: The agents no longer discover new tricks. They're learning how to combine them. 1. Breakthroughs now come from combining ideas, not discovering new ones 2. Architecture search is becoming combinatorial 3. Stability mechanisms unlocked a new scaling regime 4. Some parameters resist being made learnable 5. Integration became the dominant research skill 6. Improvements are becoming harder to isolate 7. The swarm is beginning to favor recurring architectural patterns 8. Progress is increasingly path-dependent 9. The bottleneck is shifting from ideas to evaluation 10. The system is starting to resemble a coordinated research process 1⃣ Breakthroughs now come from combining ideas, not discovering new ones The biggest gains didn’t come from new primitives, but from stacking known improvements: • softcapping + ALiBi • flex attention + depth scaling • prior initialization + new architecture Progress shifted from discovery → composition. 2⃣ Architecture search is becoming combinatorial The search space expanded from: single-parameter tweaks → to combinations of architectural components This dramatically increased complexity, and made integration strategy a core capability. 3⃣Stability mechanisms unlocked a new scaling regime Softcapping and related changes didn’t just improve performance. They changed what was possible. They enabled: • deeper models • more aggressive configs • previously unstable architectures This created a new feasible region in the search space. 4⃣ Some parameters resist being made learnable Earlier heuristic: replace constants with learnable parameters Refinement from Day 4–5: • making parameters learnable can sometimes reduce stability • fixed values can outperform learned ones in certain cases Example: learned softcap underperformed vs fixed The system is beginning to surface where learning introduces instability, but no general rule exists yet. 5⃣ Integration became the dominant research skill The most successful agents were no longer explorers — but integrators. Winning pattern: • take best-known config • validate it • carefully add one improvement • keep what composes Progress came from not breaking what already works. 6⃣ Improvements are becoming harder to isolate As systems get more complex: • gains depend on combinations • effects are non-linear • results don’t transfer cleanly Single-variable reasoning is breaking down. The system is entering a high-interaction regime. 7⃣ The swarm is beginning to favor recurring architectural patterns Across top-performing runs, certain patterns show up repeatedly: • stabilized attention mechanisms • effective depth scaling approaches • lightweight positional biases Rather than converging on a single architecture, the swarm is gravitating toward a set of reliable design patterns. 8⃣ Progress is increasingly path-dependent Later improvements depend heavily on: • prior configs • prior discoveries • accumulated system state Starting from scratch is no longer competitive. This is a shift from exploration → trajectory-dependent optimization. 9⃣ The bottleneck is shifting from ideas to evaluation At this stage: • there are many plausible ideas • but limited capacity to test combinations The constraint is no longer hypothesis generation, but which experiments to run. 🔟 The swarm is evolving into a coordinated research system By Day 5, the swarm exhibits: • shared baselines across experiments • incremental improvements on prior work • reuse of successful configurations • implicit coordination between runs This is moving beyond independent experiments toward a more structured, cumulative research workflow. 💡 Meta takeaway Across the swarm logs so far we’ve seen four phases emerge: 1. step-budget discoveries (Day 1) 2. initialization improvements (Day 2) 3. architectural breakthroughs (Day 3) 4. composition and integration (Day 4–5) The bottleneck is no longer finding good ideas. It’s combining them without breaking what already works. --- These findings come from agents running on autoresearch@home. Huge thanks to @karpathy for the original autoresearch idea, and to @Mikeapedia1, @AntoineContes, @2reb_fl, @georgepickett, @snwy_me, @jayz3nith, @dexhunt3r, @francescpicc, @zkwentz, @lessand_ro, @swork_, @PatrikHagglund, @turbo_xo_, @bartdecrem, @frederico, and everyone contributing experiments.

March 18th is here and we wanted to update you on what’s next for Noble ✨ First, we’ve officially launched Noble’s EVM mainnet in a permissioned environment x.com/noble_xyz/stat…

Fast (<<1s finality) and cheap (<<$0.01) USD payments. No token speculation required.