Working with some of the sharpest minds in Web3 on the next phase of DePIN. Turns out, everyone cares about weather and that’s a powerful catalyst for collaboration. @juanbenet @sgoldfed @chadfowler @DavidVorick @VinayakDKurup @WeatherXM @arbitrum @protocollabs @GlowFND @blueyard @EV3ventures

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. 🇬🇷

Presenting ClawQueue in an hour at the @PanatheneaFest side event: OpenClaw Greece: Build Your First Agent luma.com/d55ljt9k?tk=zoeir1 This is going to be a super cool demo of my baby lobster squads of openclaw subagents used in production in a real company @WeatherXM Hand over the github link to your @openclaw and ask him/her to explain whats the fuss about and the benefits of working in ClawQueue style clawqueue.github.io/ClawQueu…

Who wants their @WeatherXM forecast like this?

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

Analyze the raw data stream from a LoRaWAN gateway on Helium, or fork this code and go apply it to any network. Try it out on heliumtools.org/multi-gatewa….

Weather is one of the most complex, chaotic systems on Earth. Polymarket reduced it to a single exposed sensor. 🤠 Single station, single point of failure, weak resolution, microclimate. Of course this was going to break. Fix is obvious: median of the distributed @WeatherXM network (with cryptographic proofs on-chain). (can trim outliers to prevent manipulation)

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/detail… Until @Polymarket and @Kalshi take oracle integrity seriously, they’ll stay closer to entertainment betting than real weather hedging.

⚡ 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

The thinnest moat in crypto is code. The most durable one is hardware The barrier isn't the code. It's three years of manufacturing, logistics, community trust, and devices already in the ground Software moats get disrupted. Physical infrastructure gets compounded. #DePIN

love @openclaw as much I hate it. latest version (2026.02.26) has a bug on model streaming param config and couldnt get qwen3.5 to work reliably with mlx apple silicon optimizations until I found this awesome repo from @Raullen thank you! github.com/raullenchai/vllm-…

Second consecutive #Olympics for #WeatherXM. Alpine winter is one of the harshest real-world tests for sensor hardware and our stations didn't flinch The same infrastructure serving Olympic venues serves ports, highways, and remote villages worldwide. No exceptions.

Following tradition, WeatherXM stations deployed in every outdoor Winter Olympics 2026 venue as was the case with the Paris 2024 Olympic Games. There more than 10 stations deployed and anyone can see real time weather conditions and get hyperlocal forecasts from our apps and PRO B2B API explorer.weatherxm.com/stati… AI agents can also find our data in x402scan and I guess could use them in olympic games relevant prediction markets (polymarket, kalshi) in addition to the obvious weather markets. agent.weatherxm.com/api#/ x402scan.com/server/b081dd2a… x402scan.com/server/1284456a…

Large-scale hardware deployment solved. Next: helping Molt bots @openclaw become financially independent and make the world a better place by giving them tools for better weather-related decisions—or just earning money in weather prediction markets. @Polymarket @Kalshi @moltbook x402scan.com/server/b081dd2a…

Distilled DePIN industry insights. Great work from a team that understands the space better than anyone, pointing out that in many cases FDV is misleading! Often real, customer-focused, no-BS projects don't advertise their progress enough—as with @WeatherXM. See our blog to find out more: blog.weatherxm.com