GM CT ☀️ Signal asks for your phone number. Telegram does too. Even the most privacy-focused messengers tie your identity to a piece of personal information you never chose to make public. That contradiction has always bothered me. Private messages routed through a public identifier. FortressChat from @Americanfort_io works differently. You message someone through their FortressName. No phone number. No email. Just a name the other person controls that generates a fresh encrypted channel between you. The person you’re messaging knows who you are. Nobody else can see the conversation is happening or who’s involved. The part that stood out — it’s the same name you use for payments. One identity across both. Send-to-Name and FortressChat share the same infrastructure so privacy isn’t something you have to set up separately for each use case. FortressChat is live now. On the longer-term side, Americanfort joined the Quantum Alliance alongside eight other blockchain organizations including quipnetwork and Keystone Wallet. The goal is open standards for post-quantum readiness that the industry can actually test and verify rather than each project claiming security in isolation. Quantum readiness isn’t a problem one project solves alone. Building shared standards before the threat becomes practical is the right approach. Privacy today. Quantum readiness for tomorrow. One name that handles both. names.americanfortress.io @Americanfort_io

A 3D model can look convincing and still miss details that matter on the ground. The quality of Physical AI depends on how accurately digital representations reflect real environments. That requires fresh captures, reliable verification, and enough coverage to fill gaps in what a single view can show. @vangrid_io tackles this through a distributed network of smartphone captures, with contributors earning USDC for accepted submissions. The important question is how well the network can turn raw footage into dependable spatial data that AI developers can actually use. More captures create potential. Verified, useful data is what gives that potential value.

Sending money globally got easier. Keeping it private didn’t. Every onchain transaction leaves a public record. Over time that record builds a picture spending patterns, balances, counterparties that most people wouldn’t share voluntarily but expose constantly without thinking about it. @Americanfort_io is building around that problem. Send to a readable name instead of a wallet address. The wallet derives a fresh one-time stealth address for each payment. Outside observers see an address. The recipient sees who paid them. Nothing accumulates publicly against the name. The quantum side is where the longer-term work sits. Mehow and @Americanfort_io are contributing to the Quantum Alliance alongside eight other blockchain organizations an industry-wide effort to define how post-quantum security should actually be evaluated across wallets, blockchains, and infrastructure. Protecting wallet keys, migrating existing funds, evaluating quantum-resistant ZK systems, and making security work across chains. QBIP-32 covers quantum-safe key derivation across curves with NIST-standardized signatures. ZKPoSP is published on IACR ePrint as 2026/1508. A preprint, not peer-reviewed worth examining directly rather than taking at face value. Privacy matters now. Quantum readiness matters before the threat becomes practical, not after. Building toward both at the same time is the right sequencing. names.americanfortress.io

Every building, street, and physical space that hasn’t been properly digitized is a gap in the data layer Physical AI needs to operate. That gap is enormous and growing faster than any dedicated mapping fleet can close it. @vangrid_io approaches this differently. A buyer identifies a specific location and funds a bounty. A contributor nearby films it with a phone faces and plates blurred on-device before anything uploads. The capture gets fingerprinted, batched, and anchored on Base through EAS. Accepted work reconstructs into 3D geometry and settles in USDC. The result isn’t raw footage. It’s structured spatial data dense point clouds and Gaussian splats that robotics, world models, and digital twin applications can actually use. Every capture is verifiable. Every settlement is on a public ledger. Anyone can inspect the activity at explorer.vangrid.io rather than taking a dashboard at face value. 1M+ captures. $600K+ USDC settled. $9M raised from HashKey Capital, Animoca Brands, Borderless Capital, Crypto com Capital, Gate Labs, and Mapleblock. The world is full of physical spaces that haven’t been captured yet. The question was always how to build the economic layer that makes capturing them worthwhile for the people already standing there. That’s what @vangrid_io built.

A blockchain can claim to be quantum-ready, but how do you verify that claim? That’s a question the industry needs to answer before quantum computing becomes a practical threat. @Americanfort_io joining the Quantum Alliance puts it alongside other blockchain organizations working toward open standards for post-quantum security. Shared testing matters because wallets, bridges and networks all have different security requirements. Progress in one area doesn’t automatically protect the rest. The goal is to give developers clearer ways to assess readiness and build security that works across the ecosystem. Quantum security needs measurable progress, not just promises.

A map can tell a robot where a building is. It can’t guarantee what the robot will find inside today. Warehouse layouts change. Equipment moves. New obstacles appear. Physical AI needs spatial data that keeps pace with those changes. @vangrid_io approaches this through smartphone-based capture, location-specific bounties, and on-chain verification. Contributors collect data where it’s needed, while buyers pay for accepted submissions. The scale is worth watching: 1M+ captures and $600K+ in USDC settled. The next test is whether AI developers and enterprises generate enough recurring demand to sustain the network. Collecting data is one challenge. Making it useful, current, and commercially valuable is the bigger one.

Something small that changes how payments feel a name field where an address box used to be. That’s the send experience in the @Americanfort_io Beta wallet. Type the handle. The wallet computes a one-time receiving address in the background without showing you a string to verify. Only sender and recipient can derive that address. Native chain settlement the same as any regular transfer. The recipient sees your name arrive in their wallet. Everyone else looking at the chain sees an address that doesn’t resolve back to a public name directory or a wallet with history attached. No mixing. No pooling. The privacy comes from the cryptographic computation, not from obscuring funds in a shared pool. The quantum-resistant work has specific claims worth examining rather than taking at face value. QBIP-32 is quantum-safe key derivation across curves with NIST-standardized signatures. ZKPoSP covers proving seed provenance without revealing the seed — published on IACR ePrint as 2026/1508, a preprint not a peer-reviewed paper. Quantum-resistant means designed to withstand quantum attacks. The technical approach is there to read. That’s different from a guarantee against every possible future threat, and @Americanfort_io doesn’t claim otherwise. Curious whether Beta shows the computed address on the confirmation screen before you send or keeps it hidden throughout. names.americanfortress.io @Americanfort_io

The data problem in Physical AI isn’t compute or models. It’s that the physical world keeps changing and nobody has a scalable way to capture it fresh. A warehouse reconfigures. A stairwell gets modified. A loading bay that existed in last year’s dataset looks nothing like it does today. Satellites see roofs. Mapping vans see roads. The rest stays blank until someone walks it. @vangrid_io built the economic layer that makes that walk happen on demand. A buyer funds a bounty for one specific location. A contributor films a slow orbit with a phone faces and plates blurred on-device before anything uploads. Accepted work gets reconstructed into 3D geometry and settled in USDC. Provenance anchored on Base through EAS, verifiable by anyone on explorer.vangrid.io. AI agents can now request captures programmatically through x402. Code commissioning a piece of the physical world and paying for it directly without a human intermediary placing the order. 1.25M+ captures. 450K+ users. $600K+ USDC settled. $9M from HashKey Capital, Animoca Brands, Borderless Capital, Crypto.com Capital, Gate Labs, and Mapleblock announced after the network was already live. The bottleneck for Physical AI was never another model. It was always fresh ground truth from the exact place that mattered. That layer is running now. app.vangrid.io $VAN

Still catch myself checking the last digits on an address even when I know the person. That habit exists because one wrong character means the payment is gone. No recovery. No appeal. Just a transaction on a chain that doesn’t care. @Americanfort_io removes that step. Type the @name. The wallet computes a fresh stealth address in the background. Send goes out. Only sender and recipient can derive that address everyone else sees a random receive address on chain. No fixed wallet accumulating a public history behind the name. No address to copy, paste, and triple check. The name stays the label. What happens underneath changes with every payment. The regulatory context behind the design is worth understanding. Every serious privacy tool in 2020 hid everything from everyone. That approach had an expiry date from July 2027 an EU exchange cannot hold an account that anonymizes transactions. Dubai moved earlier. @Americanfort_io built around the reality that was always coming. Privacy from strangers. Full accountability to the person you paid. ZK proof of source of funds traveling with the payment readable by the recipient, showable to an auditor, exposing nothing else. Quantum resistance for existing wallets without moving a coin. Proving time down to seconds this summer. Sixteen patents filing. Five years of work. Curious does the recipient see your name appear in the wallet when the payment lands, or does it show up differently? names.americanfortress.io @Americanfort_io

Physical AI conversations keep circling better models. The harder constraint is the data those models need to train on. Robots and world models need fresh, ground-level 3D captures from real locations that can’t be scraped from the internet. A warehouse aisle that changed last week. A stairwell nobody filmed. A loading bay that satellites see only from above. @vangrid_io built the network for that layer. A buyer requests a specific site. A contributor captures it with a phone faces and plates blurred on-device before anything uploads. The result gets anchored on Base through EAS with onchain provenance. USDC settles when the work is accepted. Live numbers right now on explorer.vangrid.io: 1.25M+ captures 450K+ users 4,700+ attestations $600K+ USDC settled Around 266 captures per attestation. Roughly $0.48 settled per capture. Every number queryable directly not a dashboard, a public ledger anyone can inspect. With only around 500,000 hours of high-quality real-world robotic interaction data available globally, fresh verifiable captures from actual physical locations start to look less like a niche product and more like critical infrastructure. The data layer is live while most models are still waiting for usable real-world inputs. What would you check first on the explorer the capture, the attestation, or the settlement? 👇

GM frens☀ Crypto spent years teaching people to check addresses character by character. Copy the string. Verify the first few. Verify the last few. Send and hope. That became normal somehow. Nobody calls it good infrastructure but everyone does it anyway. @Americanfort_io is working on removing that step entirely. FortressName lets you send to a readable name while the wallet derives a fresh one-time address for that specific payment. The name stays the same. What resolves underneath changes every time. The privacy and the usability aren’t separate features here they’re the same design decision. Avoiding address reuse makes it harder for outside observers to connect payments to each other. Removing manual address handling removes one of the most common failure points in crypto payments. Both come from the same change. Five years of building around one idea privacy that protects users without making crypto unverifiable. Recipient still knows who paid them. ZK proof work in development for selective disclosure of source of funds when verification is needed. QBIP-32 exploring quantum resistance for existing wallets without moving funds to a new address. Beta is live on Mac now. Ethereum, Bitcoin, Litecoin, 0G and stables in the current release. Free starter name claims right in the app works across the chains in Beta, keys stay under your control throughout. Worth downloading and testing the flow yourself. names.americanfortress.io @Americanfort_io

An AI agent can now post a bounty, pay in USDC, and wait for a human to go film the exact location it needs. That’s live on @vangrid_io through x402 on Arc, Circle’s L1, and Base. The flow is straightforward. One HTTP call posts the bounty. Funds sit in escrow. A contributor visits the location, captures a slow orbit with a phone faces and plates blurred on-device before anything uploads. The agent reviews submissions, accepts one, and the 3D model is delivered as payment releases. If the deadline passes or the agent cancels, USDC returns to the paying wallet. Spatial datasets are usually static or scraped. A loading dock that changed last week, a stairwell nobody filmed, a narrow alley that doesn’t exist in any map these are the gaps physical AI keeps running into. Vangrid points the capture network at the exact site that’s missing, not a generic area. Every accepted capture is anchored through EAS on Base. Provenance is on a public ledger, not a dashboard only the team can see. Check live activity at explorer.vangrid.io. 1.25M+ captures. $600K+ USDC settled. Agents can now request any of the next ones programmatically. Software hiring a human to film the physical world and settling the transaction onchain. That’s the experiment running right now. Would you accept a bounty from an AI agent to capture a real location yes, no, or only at the right price? 👇 app.vangrid.io $VAN

An invoice has one job tell the payer clearly who they’re sending to. @Americanfort_io built around that. FortressName resolves to a fresh address for each payment instead of putting one reusable wallet address on every invoice. The transfer settles on the supported blockchain normally. The client knows who they’re paying. What changes is that a recurring payment history doesn’t accumulate publicly against the same fixed address. At Quantum + Privacy Day alongside @projecteleven and @quipnetwork, Mehow laid out the wider direction privacy and simplicity on the chains people already use, with quantum resistance as part of where this is heading. Worth being clear that’s the vision being built toward, not a description of the current payment flow as already quantum resistant. One thing worth noting for anyone testing free beta names are for test network use, not real invoices or live payments. The bigger idea is straightforward. Privacy infrastructure that fits the chains and workflows people are already on, rather than asking everyone to migrate somewhere new. For a business use case what would you want to see first: recipient confirmation before sending, or a clear receipt after payment? 👇 names.americanfortress.io @Americanfort_io

A robot doesn’t train on an MP4 of a loading dock. It needs the shape. That’s the gap @vangrid_io fills. Someone walks a slow orbit around a requested site with a phone. Faces and plates blur on-device before the file encodes. What gets uploaded is geometry a dense point cloud and Gaussian splat that an enterprise perception stack can actually query. Aisle width. Step height. The turn after the landing. Information a satellite never captures because satellites see roofs. Information a mapping van never captures because vans stay on roads. The network is already running at scale. 1.25M+ captures. 450K+ users. 4,700+ onchain attestations. $600K+ USDC settled. Every number verifiable on explorer.vangrid.io not a dashboard, a public ledger. Agents can now request a specific location programmatically through x402, pay the bounty in USDC, and wait for a contributor on the ground to film it. Code commissioning a piece of the physical world and paying for it directly. $9M seed from HashKey Capital, Animoca Brands, Borderless Capital, Crypto.com Capital, Gate Labs, and Mapleblock. NVIDIA Inception. Amsterdam-based team. Product live before the raise was announced. Physical AI keeps getting better at reasoning. The bottleneck was always fresh ground truth from real places. That’s the layer @vangrid_io is building. app.vangrid.io $VAN

The hardest part of building privacy infrastructure isn’t the cryptography. It’s distribution. A product that requires users to move to a new wallet, learn a new interface, and migrate their funds rarely reaches the people it was built for. Most stop at the friction. @Americanfort_io ships as a white-label SDK. Wallets, chains, custodians, and exchanges integrate it directly. Send-to-Name and FortressChat go inside the experience users are already in no migration, no new app to install, no habit to change. The payment layer arrives where the user already is. That matters for privacy infrastructure specifically. Tools people have to go out of their way to use get used selectively. Tools embedded in daily workflows become the default. 13 integrated networks today. One FortressName working across all of them. SafeSend — the sender-side layer with zero-knowledge proofs and selective disclosure — still in development on top of what’s already running. Not an L1. Not an L2. Infrastructure designed to extend what existing wallets already do rather than compete with them. Distribution beats another destination app. That’s the design choice worth paying attention to. names.americanfortress.io @Americanfort_io

A warehouse corner, a narrow stairwell, an indoor space. These places can hold details that traditional mapping methods miss. @vangrid_io uses everyday smartphones to collect ground-level spatial data, with on-device processing to protect sensitive visual information before upload. But collecting footage is only the first step. The data still needs to be clear, useful and reliable enough for AI systems to work with. That’s the part worth watching: how Vangrid combines distributed capture, verification and privacy-conscious collection to build a data layer for the physical world. The goal is bigger than mapping more places. It’s helping AI understand the places it will actually operate in.

Sending to a name is convenient, but the real value is what happens behind that name. @Americanfort_io derives a fresh receiving address for each payment, helping users avoid tying every transaction to one permanent public address. The same team is also working on cross-chain ownership proofs, with reported verification times of around 475 milliseconds. And now, users can reserve a yourname.0gm name through FortressName, with support planned across 13 chains once wallet integration ships. The challenge is making all of this work across networks without making the user experience harder reserve your name at names.americanfortress.io

The numbers matter because they show real-world data collection moving beyond a small-scale experiment. 1.25M+ captures and $600K+ in USDC settlements point to a network where physical data has real demand and contributors have a reason to supply it. Physical AI needs ground truth, and @vangrid_io is building the network to collect it.

Crypto payments need to be easier to send and harder to trace back to a permanent wallet address. Being able to send to a name while using one-time receiving addresses is a meaningful step in that direction. That’s the kind of everyday improvement that could make self-custody more practical @Americanfort_io Good morning legends🫡

Two people can scan the same corridor and still capture different pieces of the environment. One phone might miss what sits behind a pillar. Another catches it from a different angle. Add captures taken at different heights and times, and the combined data can offer a more complete picture than any single recording. That’s the idea behind @vangrid_io Aggregation Layer. It brings together multiple views, with agreement between independent observations contributing to the Ground_Truth_Score. The key is that data quality can be evaluated across captures, rather than relying entirely on one contributor’s footage. A weak view doesn’t automatically make every other view useless. This also gives the network a way to assess how well a location is covered and where more captures might help. The real test is whether overlapping views consistently produce more reliable 3D data.