Spoke at an event today for CEX traders Walked them through Hyperliquid, HIP-3, builder codes, Pear Protocol, TreadFi terminal, Based app... Watched people realize in real time: You can trade with CEX-level UX while staying the custodian of your own capital And maybe even build something on top of Hyperliquid. The future clicked for a lot of them today Job's not done though We do it one talk at a time while the team builds the House of All Finance

we were craving for Binance spot listing a year ago and here you go everything CZ touches turnes into shit lmao

Next meetup is coming. Hyperliquid.

You can't make it better then last time bro it ain't possible

dear @chameleon_jeff, i wrote you but you still aint calling, i wrote my address, my discord handle and my twitter handle at the bottom, i sent two dms to back in autumn, you must of not got em, there was probably a problem with the points system or something, but anyways whats up with season 3? i know you hear this everyday but im your biggest fan, I got room full of your posters and your pictures man, I like how you keep teasing us with farming these protocols, i hope something dope happens in singapore, anyways i hope you get this man, hit me back, just to chat truly yours, your biggest fan. P.S Hyperliquid

This is a practical accessibility example for Hermes by Nous Research. I sent a short Telegram video note and asked the agent: “Describe what is around me.” The video is muted here on purpose. The point is not my speech. The point is visual context. As a blind user, I could quickly get a description of the environment around me: I was outside, likely in a park or square. There was a wide paved walkway, grass, trees and bushes nearby. People were walking around, including children. One person was near a bicycle or scooter. No obvious dangerous obstacle was directly in front of the camera, but there was a curb and people moving nearby. That is exactly the kind of thing a blind person may want from a video message. Not a transcript. Not a generic file attachment. Not “download it and process it manually.” A video note is often voice + visual context. For accessibility, Hermes should treat Telegram video notes as first-class messages: - audio for transcription - video for visual understanding - a simple path to ask: “what is around me?” We already have a PR open for native Telegram video-note support in Hermes: github.com/NousResearch/herm… The required checks are passing. If someone at @NousResearch, @Teknium @witcheer or the Hermes maintainer team can review it, I’d be very happy to see this merged. Small feature technically. Huge difference in real use.

Despite being on the precipice of growing into a crypto “major”, $HYPE is still quite illiquid with very few holders relative to BTC SOL ETH. If you’re in this thing, either bet on the multi-year vision and ride the volatility or become an expert at trading these short term swings. Otherwise you’re going to torture yourself to death trying to micro-explain every single weekly candle. In the coming years we are going to see this token distribute across all of wall street. I have seen nothing to suggest we are off that track.

Most CT AMAs are engagement farming with extra steps. Flood’s was different because the interesting part wasn’t the alpha call. It was the operating system underneath the answers. A few bits worth bookmarking: Q: Best bear case for HYPE? A: Growth has slowed. This is not a meta where you can own things that grow slowly. The long-term dynamic between HIP-3 and Hyperliquid is still unclear. Regulation is the other obvious risk. That is probably the cleanest sober HYPE bear case from someone who is still structurally aligned with the ecosystem. Not “it goes to zero, TradeXYZ leaving.” More like: the market stops paying for slow growth, and the product surface starts creating internal tension. Q: What is the vision for fullstack? A: One deposit, trade on every exchange, portfolio margin. Maximum capital efficiency. Crypto will keep trying to build siloed venues on separate chains, which creates more fragmentation. The winning venue is the place where users can access everything without thinking about wallets or bridging. That answer is quietly the whole exchange thesis. Users do not want ideology. They want unified collateral, fast settlement, deep markets, and fewer complex steps. Q: What is the most durable edge in crypto? A: Being the most informed market participant. Not “better TA.” Not “more leverage.” Not “private groups.” Just information quality, processed faster and with less ego. Q: What are the non-negotiables for making money trading over the long term? A: Emotional control. Stable relationship dynamics. Proper sleep. Continuous research and curiosity. Low ego. The funny part is how unsexy the answer is. Everyone wants the secret trade. The actual prerequisite stack is: don’t be unstable, don’t be exhausted, don’t be arrogant, and keep learning. Q: How did you learn to trade? A: Study market history, build an edge through repeated trial and error, disappear socially for a long stretch, focus on the math and game theory, then find exceptional people who share your principles and worldview. Brutal answer, but useful. The price of “edge” is usually not one course or one model. It is obsessive repetition plus the right environment. Q: Last time you were badly wrong? A: RLB. The data turned, the business/operational issues were visible, but derisking was too slow. Cost: 400–500 BTC notional. Most of the position was sold down 60% from ATH before the exit was done. This is probably the most important answer in the whole AMA. Being early and right on huge trades does not remove the need to cut when the thesis breaks. It just makes the eventual mistake larger if you don’t. Q: Best financial decision outside picking investments? A: Living far below means and reaching escape velocity where risk-free income from T-bills exceeds lifestyle burn. That is the real “make it” threshold. Not a screenshot. Not a number in a wallet. A life where forced selling disappears. My takeaway: The useful parts of the AMA all point to the same thing. The edge is not just picking HYPE early, finding the next market structure trade, or calling the AI funding contraction before consensus. The edge is designing a life where you can stay liquid, informed, unemotional, and patient long enough to be early to the obvious trades to become obvious to everyone else. Waiting forward for the next one, Flood.

Equity correction providing the best buying opportunity of my lifetime as CAPEX spend has compressed multiples on multiple quality assets I've been priced out of

Excuse me, can someone explain why this $HYPE coin cannot just go up only. Every single day is a new emotion. One day I'm planning the beach house, the next day I'm back to eating rice. I was promised adoption, institutions, housing all of finance, S3 airdrop, $100 by EOM. This Hyperliquid thing was supposed to retire me early, not age me faster. My friends ask how I'm doing and idk what to answer because it depends on the color of some candles on a screen. Anyway I'll be checking again in 5 minutes. Hyperliquid.

I love GPT 5.6 Sol for pushing the boundaries of what is possible. This past week I've been going crazy with rewriting my harnesses, the skills and processes. So many new ideas with this new model. OpenAI truly keeps this planet spinning with the cutting edge stuff they build!

I've farmed @HypurrFi for 1+ years and all I got is this sticker. It was actually worth it.

> be Lee Robinson > Altana fund manager > short subprime in 2008 > put on $20m > walk away with roughly $200m > casual 900% > actual Big Short resume > fast forward to 2026 > see private credit everywhere > $1.8T market > opaque as hell > no clean tape > no easy short button > cannot just click “sell private credit” > so you look for the public exit wound > insurers > they love long assets > they have long liabilities > private credit fits perfectly > until it doesn’t > start buying CDS on names like Lincoln National, MetLife, Berkshire Hathaway > not because private credit has a ticker > because someone’s balance sheet has to show the cracks first > everyone watches the asset class > the real trade is finding who has to admit the losses > anyway > @tradexyz please list insurer perps > let the homies participate in Big Short Episode 2

How many of you guys would be better off if you got into coma at the moment of HYPE TGE in 2024 and woke up just today?

So Claude Fable 5 lived for three days and then started smelling like a ban. The US government put export controls on Anthropic’s Fable 5 and Mythos 5. The logic is simple: models this powerful should not be available to non-Americans. The problem is that everyone on the internet becomes “American” with one button. So Anthropic did the only thing they could do operationally: they killed access for everyone. Subscription, API, cloud providers, all of it. And I honestly do not see how this is supposed to work. Sure, you can add KYC. You can ask for passports. You can ask for selfies like exchanges do. You can make the user smile into the camera every few hours. But who is going to build serious agentic workflows on top of that? Imagine you have a long-running coding task at 3 AM and Anthropic lovingly asks you to prove that you are still the same compliant meat identity sitting in front of the webcam. Also, what exactly stops API reselling? Not even in some sophisticated “dark market” way. Any serious AI operator already has routing layers, model fallbacks, multiple provider accounts, token rotation, window management, quota smoothing, anti-expiry handling, and a single internal API that all bots and agents talk to. Once you have that, adding “one verified American resident at the front door” is not a hard engineering problem. It is policy theater meeting infrastructure reality. Anthropic is now complaining that GPT 5.5 can find security bugs across a codebase too. Well, maybe do not spend years writing essays about how AI needs to be slowed down and then act surprised when the slowdown comes for your own product, Dario. The funniest part is that everyone who touched Fable for those three days can now walk around with the achievement badge: “I had Fable running in my workflow before the empire took it away.” Claude enjoyers had about 72 hours of hope and are now back with Opus, slowly realizing that building your stack around Anthropic means spending your life rebuilding pipelines, workflows, skills, fallbacks, wrappers, and crutches for your own crutches. You pay for the subscription. They remove capabilities. They dumb down the model. They ban accounts. They change the policy. They break your workflows. Then they apologize in a blog post and call it safety. Meanwhile, a lot of serious Claude Code users have already been running GPT 5.5 inside their Claude Code workflows anyway, because otherwise the code and the workflows start looking like kindergarten craft projects. So maybe the answer is simple: Stop rebuilding your house on sand every time Anthropic changes the weather. Move properly to Codex + Hermes and stop making your own head hurt. Take the orange pill out of your brain and work like a human 2.0.

Hyperliquid retired the copytrading vault button and handed the surface to engineers. It got promoted from feature to infrastructure. So if you want to open a trading vault and start accepting deposits on Hyperliquid now, the answer is not “click the old button.” Hyperliquid priced native vault creation out of the casual UI flow by raising the cost from $100 to $10,000. In the docs, the old model is now labeled legacy HyperCore vaults: useful historically, but no longer a fit for what Hyperliquid has become. The old model was simple. A vault leader created a vault inside HyperCore. Users deposited USDC. The leader traded. Profits were split by shares. Clean PAMM-style copytrading and managed deposits, with the leader taking a fixed 10%. The problem is that legacy vaults are stuck in 2023. No spot trading. No HIP-3 assets. Weak composability. User shares never become real first-class onchain assets. For the current Hyperliquid stack, that box got too small. So the old button is not coming back in the same form. The vault mechanic is moving to HyperEVM. The new model is still permissionless, but much more technical. You launch the vault as a HyperEVM smart contract. Inside it, you can build ERC-4626 accounting, share tokens, fees, entry and exit rules, whitelists, agent permissions, risk limits, redemption logic, and your own capital-management mechanics. The link back to HyperCore runs through CoreWriter, a system contract that receives actions from HyperEVM and pushes them into the main Hyperliquid exchange. Through CoreWriter, a vault can do vaultTransfer, place orders, move assets between EVM and Core, add API wallets, and delegate authorized agents. Order actions and vault transfers carry a few seconds of onchain delay, but that is the basic architecture. So vault creation is still permissionless. It just moved from “open a vault” to “engineer a fund primitive.” You have a few real paths now. First: build your own HyperEVM + CoreWriter vault from scratch. Deploy the contract. Define share accounting, fees, withdrawal rules, execution roles, risk controls, and the strategy. Cleanest permissionless path, but dangerous if you are not serious about engineering. Second: use hyperliquid-dev/hyper-evm-lib. That is the official dev kit for HyperEVM contracts. It ships CoreWriterLib, bridging helpers, precompile reads, a token registry, and useful contract functions. Same path, less low-level plumbing. Third: study open-source templates like manhashed/hyperliquidvault or adamocallaghan/hyperevm-vault. They give you pieces: CoreWriter integration, deposit and withdrawal flows, limit orders, bridge logic, API wallet support, and ERC-4626-style vault structure. But these are not deploy-and-start-collecting-deposits products. They are scaffolds. Before touching real capital, you need tests, threat modeling, permission review, withdrawal logic review, role review, and a proper audit. A vault is not just a contract. It is custody, accounting, execution, permissions, and reputation bundled into one failure surface. Markets are hard enough already. You do not want extra blow-up risk from sloppy vault infrastructure. Is anything ready out of the box? Hypermax SmartFund looks closest to a productized managed-fund wrapper on top of HyperEVM: ERC-4626 share tokens, custom fees, flexible redemption periods, HyperCore liquidity and orderbook access, and risk controls. Promising, but still early. Not institutionally de-risked yet. D2, Rip, and Raga show vault-like products are live in the ecosystem. But none gives every outside manager a self-serve button to open their own vault, so they are not replacements for the old native flow. Old HyperCore vaults were a product. New HyperEVM vaults are a builder surface. The old UI button is gone. The vaults came back as infrastructure. The de facto top spot in this new vault business is still wide open. And Jeff already handed you the building blocks.

Recordered a podcast on Hyperliquid with Zerk and Trendy, dropped some alpha on AI agentic trading on Hyperliquid and how it's different from the generally horrendous trading bot experience you've seen so far. Thanks for hosting me guys, waiting forward for the Episode to drop on Spotify/Apple!

Ask not what Hyperliquid can do for you, ask what is that that you can do for Hyperliquid