WeatherXM

🚀 We’re LIVE on Amazon To kick things off, we dropped our WiFi station to €94.99 for the first wave This week = push the algorithm + grow the network fast Limited units at this price — once gone, it’s over 🇫🇷 amazon.fr/dp/B0FVDKCG1T 🇮🇹 amazon.it/dp/B0FVDKCG1T 🇪🇸 amazon.es/dp/B0FVDKCG1T

Athens had a moment last week. I spent three days at @PanatheneaFest 2026, and the takeaway is simple: Athens is now on the global tech map, a place serious people choose rather than tolerate. Panathēnea is a not-for-profit, run almost entirely by students and recent graduates. That volunteer team still pulled in founders from ElevenLabs, Bolt, Runway, Airwallex, Deel and Plum; partners from Sequoia, Index, Atomico, Balderton, Northzone, 500 Global, PayPal Ventures and Dawn; operators from OpenAI, NVIDIA, Google, Microsoft and Qualcomm; and European deeptech founders from Proxima Fusion, Isar Aerospace and Open Cosmos. It also brought Greeks home, @real_ioannis (Reflection, ex-DeepMind) and @agermanidis (Runway). And teams are putting down roots here. @dionyziz' @CommonPrefix (~40 people), @poddotnetwork, and @Mysten_Labs Sui hub on the crypto-infra side; @YSmaragdakis' Dedaub on smart-contract security. Home-grown scaleups are real too: Viva.com (Greece's payments unicorn), Workable, Blueground, agritech-AI Augmenta (and more). I spoke at the Common Prefix × Pod side event on "Blockchain and Trust Technologies," alongside @sagrawal (Pod) and @nikil511, CEO of @WeatherXM. We got into the AI × crypto thread that's stopped being hand-wavy: decentralized training. You no longer need everyone in the same data center to train a frontier model. ~99% of a model's weights are bit-identical between training steps, so you ship only the ~1% that changed. @Hevalon's team at Covenant formalized this (their PULSE paper) and pre-trained Covenant-72B with trustless peers over the open internet. @huggingface then shipped the same trick in the open. Crypto's coordination layer + a ~100× smaller payload = a big model trained across machines no single entity controls. The other shift I keep noticing: most people still use agents for coding, or as a Google alternative. What's coming is general-purpose agents that research, draft, plan and do your work for you, and the hardware already knows it. The silicon being built isn't sized for one assistant; it's sized for an agent fleet. Two days ago NVIDIA announced the DGX Station for Windows: a GB300 ARM desktop with up to 748GB of unified memory, built to run hundreds of concurrent agents on-prem. You don't build that for a chatbot you visit. If you don't have 5, 10, 50 agents running at all times, local and cloud, doing evals, auto-research, quietly improving work you've already done, you're probably not using AI properly yet. We're early. But the people building the hardware have placed their bet. To the Panathēnea team: thank you, and bravo. See you in 2027. 🇬🇷

The rollout of the stations of Beaver NFTs is behind schedule. We are working on a new platform to coordinate with deployers on very specific station locations and this makes everything more challenging and slow. We want to avoid vague placement of stations for the sake of rewards. Your support via the NFT campaign is much appriciated and will eventually unlock fair rewards and services for you.

@WeatherXM Whats happening for those of us that purchased the Beaver NFT’s for new deployments? I purchased 4 months ago and have not heard anything about these deployment and have not been allocated any rewards for holding these NFT’s…. It was a $400 investment???

WeatherXM is entering a new phase. As one of the earlier DePIN networks, we’ve faced many of the hard problems first: useful coverage, targeted deployment, sustainable rewards, real-world demand, and coordinating physical infrastructure at scale. 1) Now we’re evolving. We are redesigning WXM tokenomics and the WeatherXM reward mechanism. 2) At the same time, we’re refactoring core infrastructure for the next generation of WeatherXM services, including AI-powered products. Some short-term turbulence may occur across rewards, the PRO API, and mobile apps as we upgrade the system. 3) We are also returning to our foundations: DEX-first, more open source, and stronger real-world utility. WXM will no longer be listed on Gate, while DEX pools remain operational. 4) We’re also preparing new WeatherXM hardware with open firmware, including mesh / Meshtastic-compatible capabilities. To prepare for this next generation, we’ll launch a discount campaign for current-generation stations while stock lasts. The first era proved a community can build a global weather network. The next era is about making it more useful, open, resilient, and sustainable. blog.weatherxm.com/weatherxm-comm…

Momir, moat will not be at the AI / agent level but at the data level, once we have an AI trust protocol that agentic harness can consume. In many ways, DePIN projects like @WeatherXM are building this protocol without realizing it. The trust harness for physical-world AI: a blockchain-native spatial intelligence layer that lets agents safely act on hyperlocal reality, not just digital information. btw, I came here to inform you there is an impersonator of your team scamming people x.com/jiaping_vc He sends fake NDA google docs that need to install stuff on local machines :-/

We're excited to share the speaker lineup for Blockchain and Trust Technologies, our side event at Panathēnea 2026, hosted together with @poddotnetwork.

ClawQueue is live 🦾 This is something I've build and use in @WeatherXM last few months to have @openclaw write @github issues for me and turn them in AI work queues that a scheduler dispatches on local machines, while team can follow up via project boards. CQ is intentionally small: GitHub holds the durable work contract, OpenClaw helps shape the work, your machine runs the workers localy, and workflow policy stays in markdown/config you can tune with your lobster. This is a good idea, if you operate your own company/project with your own profile, agents, boards, and worklog - or - you wanna contribute to an external/open-source project through. Ask openclaw to install CQ and create a project-specific CQ profile from the upstream repo’s README/docs/contribution rules, then routing issue-driven agent work into reviewed PRs Try it: clawqueue.github.io/ClawQueue/

Strange how you can not read public numbers in our explorer: explorer.weatherxm.com 56 active stations in Kenya sending real-time data now! +2k active stations in US, +6k active globally, out of +9k deployed. Yes, 3k stations are offline, and many are of the targeted rollout deployments in developing nations that deployers came across a ton of problems unknown to the western world. e.g. No availability of AA batteries, no internet connectivity, etc. Thats the reality of the physical projects! We managed to put stations in places that had nothing before and this data is made available realtime to governments and private weather companies, used daily to improve forecasts, via partners like: synopticdata.com/blog/decentral… So please, do not bring your hate for the crypto bs projects, to a depin project that has created so much value to the world.

@WeatherXM @_veri_fi @BLCKIoT @Princeton @helium Strange you are advertising this, since most of the Swissborg deployments are now offline. Almost every station in the America's is dead, and most of those in Kenya too. Certainly not free for the investors who got nothing in return.

Joni @_veri_fi from @BLCKIoT presenting at @Princeton decenter the amazing work they did in Kenya, deploying +100 of our @helium powered weather stations, using subsidized (free) hardware part of our "targeted rollouts" campaigns. #DePIN

@ntaschetti we ship globably from our eshop! Amazon is an extra distribution channel for us. weatherxm.com/shop/

@WeatherXM 🇦🇷 ?? lamentablemente estamos muy al fin del mundo y mucha burocracia, pero espero paciente, ya me tocara el turno Saludos!

🚀 We’re LIVE on Amazon To kick things off, we dropped our WiFi station to €94.99 for the first wave This week = push the algorithm + grow the network fast Limited units at this price — once gone, it’s over 🇫🇷 amazon.fr/dp/B0FVDKCG1T 🇮🇹 amazon.it/dp/B0FVDKCG1T 🇪🇸 amazon.es/dp/B0FVDKCG1T

Last week Météo-France filed a criminal complaint. The target: whoever held a hair dryer to a weather sensor at Charles de Gaulle airport. Twice in April, that sensor briefly read 22°C in an 18°C afternoon. Twice, a Polymarket user had wagered exactly these numbers. Payouts: $14,000 and $20,000. Polymarket switched its Paris settlement to Le Bourget and refused refunds. CNN, Le Monde, Reuters and Euronews all ran the story. The hair dryer is the funny part. The architectural failure underneath is what nobody is talking about clearly enough. This is not a Polymarket scandal. It is a weather-data scandal. The bets are how we found out. One Météo-France sensor at Roissy was both the official Paris temperature and Polymarket's settlement source. No cryptographic signature. No real-time cross-check. Mains power, network connection, admin console accessible to staff. The trust model collapsed at every seam but until recently, no one had a financial reason to attack a thermometer. The hair dryer is the lowest-cost version of this attack. As our CEO pointed out an insider with admin access at any met agency could alter data without ever touching a sensor. Hair dryers are loud. Database queries are not. x.com/nikil511/statu… Weather data has stopped being just a public good. It is now the settlement layer for an economy: parametric insurance, energy contracts, agricultural payouts, emergency response, prediction markets. The hair dryer exposed a $34,000 mistake. The same architecture moves decisions a thousand times that size. This is the failure mode WeatherXM was built to remove. Every reading is signed at source by a secure element on the device. A hair dryer can still warm a sensor but it cannot forge a signature. We never settle anything important on a single station. 9,600+ stations across 98 countries, in clusters and verifiable maths, never single points of trust. Every reading is cross-checked against neighbouring stations and third-party data. Anomalies get flagged or discarded by our quality of data filters (QoD). Our WiFi & Celular stations are energy autonomous, solar powered. No mains plug, super easy to deploy. Our H2 station takes it one step further, using @helium IoT / LoRaWAN operates with decentralized wireless infrastructure that the community or customers can expand on demand to cover areas that lack celular coverage. Data is on decentralized file systems @IPFS @Filecoin @AkaveCloud and QoD based rewards are blockchain merkle-trees anyone can audit. The fix is not to regulate prediction markets harder. It is to stop trusting infrastructure that was never designed for adversaries. Build the verification into the data itself and adopt decentralized, trusteless approaches in the traditional weather industry too, as it seems we are gonna need it more than ever in the comming future. The world has changed. The traditional weather data layer needs to change too!

This issue goes far beyond prediction markets. Today’s forecasting systems depend on airport weather observations, yet much of that infrastructure was never designed for adversarial settings. An insider in a meteorological organization (e.g. admin) could alter data to influence an outcome, without a hairdryer on a sensor, and probably without being detected. We need to transition to forecasting systems backed by denser, lower-cost, and more resilient observation networks. That is exactly what we’re building at @WeatherXM robust, blockchain-enabled weather stations and verifiable weather data pipelines for the next generation of forecasting infrastructure.

@alvarogracia @nikil511 x.com/WeatherXM/stat…

Wouldnt happen with @WeatherXM @nikil511

@0xGeebz @aaronjmars x.com/WeatherXM/stat…

@aaronjmars @WeatherXM can fix this

holy fuck, a hair dryer at a Paris airport broke Polymarket weather markets & made someone $34,000 richer - polymarket was settling Paris temperature bets on a single Météo France sensor sitting near the Charles de Gaulle runway perimeter - basically unguarded - the guy bought the long-shot outcome (like "22°C" when everyone expected 18°C) for pennies, since nobody thought it'd hit - then he walked up to the probe and briefly heated the air around it with a portable heat source, spiking the reading just long enough to register as the daily max - temperature snapped back to normal in minutes, the market resolved in his favor, and he cashed out - twice, on April 6 and April 15, before Météo France caught on and filed charges hyperstitions.

@takisoul @deanmachine x.com/WeatherXM/stat…

@deanmachine Ping @WeatherXM I have an idea

Grab a hair dryer, go down to your local weather sensor, heat up the competition

We’ve already shown a better way to build a weather oracle: median temperature across dozens of our stations, with cryptographic proofs on-chain. markets.weatherxm.com/details/london Until @Polymarket and @Kalshi take oracle integrity seriously, they’ll stay closer to entertainment betting than real weather hedging.

Every crypto whitepaper for the last decade has warned about the oracle problem. Someone finally demonstrated it for $34,000 using a hair dryer. Polymarket settles Paris temperature bets on a single Météo France weather station at Charles de Gaulle. Louvered enclosure on a pole, thermometer inside, no fence, no camera, on the perimeter of a runway that anyone can walk to. The blockchain stack resolving this market represents years of engineering. The sensor is a piece of airport equipment in a plastic box. The attack runs on a structural feature of weather reporting. Stations report the daily MAX, not the average. One brief spike sets the official record for the full day. The long-shot outcomes, like "22°C when consensus is 18°C," trade for pennies because nobody expects them to hit. Buy thousands of shares of the improbable outcome for $100. Walk to the sensor. Heat the air near the probe until the reading crosses your strike. Cash out. He did it twice. April 6 at 6:30 PM, the sensor jumped 4°C in 12 minutes and then collapsed back to normal. April 15 at 9:30 PM, same pattern under calm cloudy skies. Neighboring stations showed nothing. Wind and humidity flatlined through both events. The data forensics read like a confession. Polymarket paid $14,000 on the first one and $20,000 on the second. One of the winning accounts was created days before the first anomaly. Polymarket's response was to change the resolution source from CDG to Paris-Le Bourget. Another public weather station down the road, same security posture. The oracle problem is unchanged. They picked a different single point of failure. Every Polymarket weather market has this shape. Every sports market resolves on a single stats provider. Every election market reads from a single wire service. The attack pattern is identical across all of them: find the weakest reporter, influence the reading, collect the asymmetric payoff. This one required a walk to a Paris runway. Others require a phone call to a data entry clerk. Météo France filed the criminal complaint, not Polymarket. The blockchain executed exactly what it was designed to execute: a deterministic payout on a corrupted input it had no way to verify. Garbage in, garbage out, cryptographically signed. The hardest part of crypto is the chain. The weakest part is the thermometer.

⚡ Day 2 on Amazon — and the response has been insane Huge thanks to everyone who already grabbed a station 🙌 Reminder: launch price is still €94.99 for the first batch Once these units are gone, price goes back up If you’re still on the fence, now’s the time 🇫🇷 amazon.fr/dp/B0FVDKCG1T 🇮🇹 amazon.it/dp/B0FVDKCG1T 🇪🇸 amazon.es/dp/B0FVDKCG1T

@streeter2203 weatherxm.com/discord

@WeatherXM Yes 👀

We have setup up our own markets, resolving using high QoD median of 50+ stations with data proofs on chain. markets.weatherxm.com/overview Problem is bootstrapping liquidity. So weather underground as oracle and multiple insecure layers to tamper data, are here to stay. Thats the web2 mechanics of web3 prediction markets. Anyway... here is a recipe to win the existing weather prediction markets: Create an ensemble virtual station, of near by stations, use historical observations to find the best fit and apply bias correction on combo if needed. Benefit from faster updates. Create the same ensemble of MOSAIC forecasts (WXMv2) which are basically MOS forecasts, benefit from multi model, multi station virtual hyperlocal forecast, which handles edge cases better. ...and ask your friends to deploy more stations in the areas that Polymarket & Kalshi resolution stations are so we can all benefit before (more) people figure this out. We have huge promo discounts on our eshop (shipping globally) and Amazon (DE,FR,IT,ES) x.com/i/status/20462…

@jhiller @GrablyR @helium @WeatherXM i genuinely think we’ve got edge here

Et si des parieurs de Polymarket avaient manipulé... des capteurs météo? Auprès de BFMTV, Météo France confirme avoir déposé une plainte suite à des traces d'accès illégitime à leur capteur de température de l'aéroport Charles de Gaulle. Le 6 avril dernier, il indiquait subitement une température maximale (très inattendue) de 21 degrés aux alentours de 19h. Faisant gagner 14.000 dollars à un internaute qui venait de créer son compte.

@streeter2203 Uh?

@WeatherXM Station broken and not able to join discord