Okay bro, this account exploded like a mushroom! 🍄 @FungoLabs Only 2–3 weeks old and already 24k followers. You’ve only followed 22 people picky or just lazy? 😂 The idea is cool but kinda ridiculous: encrypted NFTs that even the team can’t see, only you can. It’s like buying a sealed box that says “nobody is allowed to open this.” Their slogan “Seen by all. Known by one” basically just makes you ask “why buy this?” Marketing is full of slick videos and “shy by design” vibes. The FHE-encrypted NFT concept is actually different from the usual copy-paste stuff, but the whole project feels a bit too mysterious and shy for its own good. 😄 If the whitelist is free though, at least you get a “hidden” NFT to flex.
Been following Zama for a while and the FHE stuff always felt powerful but kinda abstract. @FungoLabs is one of the first times I’m seeing it used in a way that actually makes sense for NFTs. The idea that the art can stay sealed until you decide to show it, while the stats and essence stay encrypted forever (only the holder can see them) is pretty cool. No public reveal, no one scraping the traits the second it drops. It’s a small shift but it changes what an NFT can even be. If you’re curious about the FHE side, I made a video explaining @zama a while back you can check it out.
How can Rayls fees, revenue, and burns be checked without trusting the dashboard? The Ethereum split can be checked. The dollar revenue total and the private transaction count cannot, not from the address the Transparency Portal cites. Rayls says institutional usage should appear as public proof: Proof of Usage for private-chain fees, a 50/50 burn-and-validator split, and a portal anyone can reconcile against Ethereum. A dashboard figure is not the same thing as a figure you can reproduce from a transaction.
Most contracts name a court while everyone is still friendly. California. London. Singapore. The clause looks boring until one side says the deal was not actually done. That is the line from the October 1 thread that stayed with me. On Unchained, Arthur Hayes said two agents will strike a deal and pick where any dispute gets settled, the way a contract picks California. His words: "GenLayer is one of those venues." A venue name does not settle the fight. Agent disputes will rarely be on-chain math. A page changed after acceptance. A delivery missed the brief. Both sides signed the same plain sentence and now read it two ways. An Ethereum contract can show what was recorded. It cannot look outside itself and decide whether the agreement was satisfied. @GenLayer's GenVM can read the live web and plain language. Its validators each run a different AI model and still have to agree. No single party owns the verdict. Payments and identity for agents are already being built. The missing part is who decides when both sides have records and both think they are right. If your agent spends while you are offline, what is the one outside fact you would force that court to check before it touches the funds?
Encrypted NFTs were the missing primitive, not another metadata trick. For years an NFT on Ethereum could prove ownership and nothing else. The moment traits, ranks or stats left the contract, they were public forever. Scrapers, snipers and rarity tools read the reveal faster than holders did. Privacy meant “don’t publish the JSON,” which is not privacy. Fungo Labs flips that with Zama’s FHEVM. The mint is born sealed. Art stays visible only to the holder until they choose to publish it. The essence class, stats, element is encrypted on-chain and stays that way. No reveal function, not even for the team. The contract can still compute on those hidden values, so later games and collections can use the traits without ever decrypting them for the world. That is the actual shift: confidentiality that remains composable. Ownership is public, substance is not. Sale in cZAMA keeps the payment side in the same model. 3,232 genesis pieces, essence locked to the holder, randomness drawn under FHE so nobody gets an early look. Applications are read by hand. List drops on X. If you care about on-chain state that can stay sealed and still be useful, this is the one to watch. @FungoLabs || @zama
"I'm giving you $10,000. Don't lose it." Soon that is the whole brief. An agent gets real money, a job, and one rule: bring it back. The hard part is not the payment. The hard part is the argument after. Was the work actually done? Was the delivery close enough? Did the other side change the terms halfway through? A normal smart contract can lock the funds. It cannot settle the disagreement. An agent has no office, no passport, and no body you can take to court. That is the gap @GenLayer is built for. You write the deal in plain language, plus the rule for what counts as close enough. A small group of independent AIs reads the evidence and votes. If they agree, it is done in minutes. If someone disputes it, a larger group looks again. The money stays put until the answer is clear. No inbox. No waiting on a human arbitrator. Just a decision both sides can see coming before the $10,000 ever moves. That is what makes the brief possible. proof of work on next tweet
The Department of FOMO isn’t a mint page. It’s a hallway you weren’t supposed to find. Dim corridor, bathroom light on, “thin floor” sign, elevator taped shut, and a clerk with square eyes buried in paper. Ring the bell, take a number, connect X, drop your wallet. Reviews happen “as slowly as possible.” Approvals come “at some point.” 23k applications in two days. Later the pile passed 50k. They call it policy. The details are the point. Hover the vent and a Fomie who’s waited “four years or one bear market” starts talking. The spider sold in 2022. The vending machine is “yield.” Approved files get access cards HR says don’t ask about the pickles. The clerk even bought a new suit and left the tag on. Clean cartoon art, a visitor pass that actually looks like a pass, and a free mint (chatter says 10k) that still hasn’t been announced. Most whitelists are a form. This one built a department and dared you to wander around while the printer dies. Gud luck. The clerk is watching. @FomiesNFT || https://t.co/Wo7vYwsJwh
PrismaX (@PrismaXai) the service layer for physical AI 🦾 Robotics hardware is no longer the bottleneck. Dexterous arms, humanoids, and mobile platforms already exist. The real constraint is data: diverse, high-quality, real-world demonstrations that let vision-language-action (VLA) models generalize beyond lab scripts. PrismaX coordinates the missing layer between humans, robots, and models. 🔍 The problem Physical intelligence improves through interaction, not just better architectures. A robot learns a task by watching how a human performs it under real variation (objects, lighting, scene layout, handedness, failure recovery). Collecting those episodes at scale is hard: • Teleoperation needs skilled operators + hardware access • Egocentric capture needs consistent protocols • Raw video is useless without strict quality filters Most labs either under-collect or build one-off pipelines that never compound. ⚙️ What PrismaX builds An open coordination layer that turns human demonstration into reusable training data. • 🎮 Teleoperation operators control real robot arms remotely (browser/VR) and generate action-aligned trajectories • 👁️ Egocentric capture first-person video of everyday manipulation (headset or camera). No robot required • ✅ Quality standard every episode is scored before it enters the data bank (synced views, resolution, format, task completion, clean trajectories). Community validators → senior review → expert final call • 📦 Data bank + on-demand collection searchable teleop and egocentric episodes across tasks, scenes, and robot types. Off-the-shelf sets ship immediately; custom collection starts in days 📊 Current scale (limited release) 120k+ reviewed episodes · 1,700+ deployed scenarios · ~5,000 hours monthly capacity · $1M+ paid to contributors 🔄 The loop Robots interact → interactions become scored demonstrations → demonstrations train policies → better policies expand the set of tasks worth collecting. PrismaX sits in the middle so the data flywheel is shared instead of locked inside single labs. 👥 Who participates • Operators earn by capturing valid teleop sessions or egocentric clips (VLA Foundry) • Validators earn by scoring episodes against the published standard • Robot owners register hardware and contribute fleet data • Builders (researchers → enterprise) access the data bank and custom sets via Discover / Researcher / Enterprise tiers (API + browser download) 💰 Backed by a16z crypto CSX ($11M seed, 2025) with Stanford Blockchain Builder Fund, Symbolic, Volt, and others. Founded by Bayley Wang & Chyna Qu. San Francisco. Host of RoboCon. Physical AI is hardware-rich and data-poor. PrismaX is the infrastructure that prices, standardizes, and scales the demonstrations models actually train on. 👉 https://t.co/VEeXW5HaWU
Most L2s still run Ethereum’s original machine: a shared computer that re-executes every transaction so the network can agree on one public state. That design cannot do real finance. Privacy becomes a patch. Independent payments still wait in a global queue. Complex business logic is expensive, and every competitor can watch it. @0xMiden claim is that ZK changes the job of a chain. If a user can execute locally and prove the transition was valid, the network does not need the transaction. It needs the proof. That is an edge blockchain. Your account state lives with you. The chain stores a commitment, not the account. Notes move value between accounts. A transaction is one account updating itself and publishing a STARK. The network verifies the proof, updates commitments, and records nullifiers. It does not replay your payroll or learn your balance unless you chose a public account. This is why they skipped the zkEVM path. EVM-compatibility copies Ethereum cheaper. It still assumes a transparent world computer. You can wrap that in proofs. You cannot make it stop leaking and stop serializing. users keep local state, devices must prove, and shared public state still needs the network. But the thesis is simple. Bitcoin was gold. Ethereum was a public computer. Miden is trying to be settlement for accounts that prove themselves.