This is what September looked like for @axisrobotics A month where conviction met proof. We crossed 5M robot data trajectories on @base and we’re already past 6M, growing every day. 200K+ contributors worldwide are helping us build the physical experience layer for robots. Our Franka dataset family crossed 160K+ downloads on huggingface, becoming the most downloaded open-source simulated Franka manipulation dataset. We sat down with brian_armstrong during his visit to Singapore to talk about the future of Physical AI and where robotics is heading. Huge thanks to baseapac for making it happen. We also joined BasedIndia to share our approach: building a scalable human in the loop system where anyone can contribute to robot intelligence. Our research was accepted to the IROS PWMS workshop another step toward scaling embodied AI through diverse, real world inspired data. And then came one of the biggest community moments of the month. 2,472 participants committed 2,394,209.52 USDC in the Axis Community Sale on Sonar. The most participated Sonar sale of 2026. To everyone who participated, contributed, downloaded our datasets, tested our work, or believed in the vision thank you. We also announced 25% TGE unlock for everyone, plus additional fully unlocked bonus tokens based on original commitment. September wasn’t just about milestones. It was about proving that the community, data, research, and infrastructure can compound together. 6M+ trajectories and counting. The engine keeps running. Robotics, Based.

Big milestone for @primus_labs Primus has successfully raised $6.5M across its Pre Seed and Seed rounds a strong vote of confidence from some of the biggest names across crypto, Web3, and technology. The round was led by @DispersionVC, @symbolicvc, and @vaneck_us, with support from @SamsungNext, @Alchemy, @MaelstromFund, @welaraworld, @theBBFund, @Modular_Capital, @Karatage_, @InitialVentures, @Mozaik_Capital, and @ArweaveSCP. Huge appreciation as well to Legend Star, @ambergroup_io, @snzholding, @HashGlobal, and the incredible angels supporting Primus from the beginning. Raising capital is only one part of the journey. The real opportunity now is turning that momentum into products, adoption, and infrastructure that can create lasting value for the ecosystem. $6.5M is a major milestone, but it feels more like the beginning than the destination. Congrats to the entire Primus team and everyone who believed in the vision early. The next chapter is going to be exciting. 🔥 Onward and upward! 🚀

Most spatial data is sent raw to the cloud, where faces and license plates are filtered, if they are filtered at all. @vangrid_io reverses that order. The pipeline has five stages: capture, edge compute, cryptographic signing, aggregation, then API delivery. At the edge, faces and license plates are blurred on the phone before the video is encoded. Frames that remain readable are never uploaded, never stored, and cannot be recovered. Provenance is attached at the source, rather than being a promise made after the data reaches the server. Anchored data can be accessed by machines without an account or API key. The client calls data.vangrid.io/api/v1. The server responds with 402, quoting prices in USDC on Base or Arc. The client signs the payment and retries the request. Settlement gas is covered by the service. POST /spatial/query is quoted at $0.01, GET /observations/{id} at $0.005, and prices can change because they are requoted on every request. GET /provenance/{hash} is free. Unknown hashes still return 200 with verified: false. When checked, the board showed $589,659 settled, $985 still in escrow, 893 bounties, and 14 bulk orders.

Good Morning Quip Network Something I find particularly important about @quipnetwork is its approach to the security of digital assets, especially when people start discussing the long-term impact of quantum computing Most blockchains depend on cryptographic systems that were designed around the computing capabilities available today A sufficiently powerful future quantum computer could threaten some of the cryptographic assumptions behind digital signatures, which makes preparation an important topic for the entire industry @quipnetwork is working on post-quantum protection through hash-based signatures and additional wallet security mechanisms designed to protect assets across existing blockchain ecosystems What I like about this approach is the focus on practical compatibility Users and developers already have established networks, wallets and applications, so introducing stronger security without ignoring the existing ecosystem is a meaningful challenge, there is still a lot of research, testing and engineering ahead, but I would rather see projects explore these risks early than wait until quantum hardware becomes a serious threat Security infrastructure is not always the most visible part of Web3, yet it could become one of the most important

Jumper allocation secured on Legion. → $48.6M committed → $3M hard cap, over 16x subscribed → 9,690 applications Got $1,379 accepted out of my $3K. Wanted more, but with demand like that, I’ll take my seat. We jump.

You can literally make a game by describing it. You tell @RemixGG_ what you want, and it builds a playable version. I just found out about this and I’m testing it properly this week. I’ll show you what I build, what actually works, and how the Ambassador rewards work along the way. Let’s see how far this can go.

Perp or future trade - what should we know before entry to trade? - take your decision - setup your trade amount - set your leverage ( less leverage less profit, big lev. big pnl { red or green}) - choose one ( cross or isolated) - support, resistance, trading heatmap , money management and risk management. - set stop loss and take profit - ETF chart > Useful tools: - tradingview > chart - coinglass > Funding OI Long/short Liquidation -bookmap > order book - CPI, FOMC > news -trade journal app > Journal Avoid: - don't copy signal bot or group - revenge trade - blind trade with emotion - start trade without knowledge If I mistake you should share info, I want to learn more and more.

are you a perp trader, or are you looking to start trading perps? i want you to know that the perp market is now live on @Polymarket. polymarket has made perp trading simple. you can start with as little as an $8 deposit and enter a trade with as little as $0.30. whether you’re a new or experienced trader, you can’t afford to fade polymarket right now. with the token coming soon, you can also farm the @Polymarket leaderboard for potential airdrop rewards while you trade. start here: polymarket.com?via=ifycalex signup with your gmail account and make deposit. you can also predict outcomes across crypto, sports, politics, markets, and other real-world events, all in one place.

Halloween is getting closer and @1win is already bringing the spooky vibes 🎃 Haunted Treats by Endorphina is now live with a massive €250,000 prize pool. From zombies and witches to vampires and some serious prizes, this one definitely has the right Halloween energy. If spooky games are your thing, you might want to check this one out. Who is going for the treats? 👀

I was checking out the latest from @PlayOnMint and noticed the MintABear update. There are 10 WL spots for the upcoming Free Mint on Robinhood Chain. I checked the details, and it’s actually pretty simple. No complicated tasks or requirements. Just join the Discord and wait for the selection within 72 hours. If you’re interested: discord.gg/mintio Let’s see who gets one. @PlayOnMint

Omg, this is so freaking unfair and not nice. I can't believe they did that, wtf... stealing ideas is so not cool, freaking insane At least it proves you built something right though bro.

most panels talk about access like it's still a pitch. ORO's already doing the unglamorous part: you tell it what you want, it moves the capital, you just sign. ZIGChain, Ethereum, Solana, and more chains on the way. that's the "distribution that reaches everywhere" line, except it's a message instead of a keynote. deFi still feels complicated to a lot of people. the product that wins is the one that makes it feel like a conversation.

I used to think losing money was the expensive part. Turns out, sometimes the really expensive part is proving that you were right. Imagine two AI agents making a deal. One says: I did exactly what we agreed. The other says: No, you didn’t. The transaction already happened. The money already moved. So who decides what actually happened? That question becomes a lot more serious when you look at the story behind @GenLayer. A custodian once lost access to around $150M because of a key issue. Getting the assets back turned into a legal battle. The reported legal cost reached roughly $1M before the whole process was over. And what remained? $14,000. That number stuck with me more than the $150M. Because @kstellana looked at a situation like that differently. Not simply as a custody failure, but as a design problem: If resolving a dispute costs more than the thing you're fighting over, the system itself becomes part of the problem. So why not build something different? That is where @GenLayer gets interesting. The idea is basically an internet native court for a world where AI agents can make agreements, interact with each other, access external information, and make decisions at machine speed. But instead of one person deciding who is right, disputes can be evaluated by a decentralized network of validators. Different validators can use different AI models. They examine the evidence, interpret the agreement and reach a collective decision. No single party gets to simply say: Trust me, my version is correct. And I think this becomes much more important as AI agents become more autonomous. Today, a disagreement between two people might take days of messages to resolve. Tomorrow, two agents could execute a transaction in seconds and disagree about the outcome immediately. The transaction is fast. The dispute isn't. That mismatch is the real problem. GenLayer isn't trying to make disagreements disappear. It is trying to give them somewhere to go. Because if AI agents are eventually trusted with serious economic decisions, we need more than fast execution. We need a credible way to answer one simple question: What actually happened? And maybe that is the part of GenLayer I find most interesting. Not the trillion dollar predictions. Not the hype around autonomous agents. But the possibility of having a dispute resolution layer built for the internet itself. Because when two sides insist they're right, someone still has to check the rules and the evidence.

Cursor makes email marketing so easy One prompt and all is done with @nitrosendx cursor

RT @ant1grav: Fortune of Olympus on @duelbits. $1k. no speech.

one detail about @agenticscredit I had not paid much attention to the first credit venue is Avantis on Base that matters more than it sounds the agent does not just get capital and disappear with it credit is routed into an approved venue where the rules and risk limits stay enforceable feels like the credit layer is slowly becoming real infrastructure not just a score on a dashboard

RT @AB_geniuss: Most people are still writing about Hazels like the mint is the whole story I spent the morning inside Studio instead and t…

Axis September Wrapped: Conviction Meets Proof Redefining the landscape of Physical AI and robotics, @axisrobotics achieved massive milestones across its ecosystem throughout September. Here are the key highlights 👇 🚀 Key Highlights and Milestones Scaling the Data Engine on Base Crossed 5 Million robot data trajectories on Base! With over 200K+ contributors worldwide, the engine continues to expand every day. Featured Sit-Down with Brian Armstrong: Axis CEO @chris_anm0 sat down with Coinbase Founder Brian Armstrong (@brian_armstrong) to discuss how Physical AI and on-chain robotics will transform the world. Most Participated Sonar Sale of 2026 In the Axis Community Sale on Sonar, 2,472 participants committed 2,394,209 USDC (2.39x oversubscribed) in just one week, demonstrating overwhelming community conviction. Academic Recognition and IROS 2026 The paper "AXIS: A Growable Community-Driven Data Engine for Scalable Robot Manipulation" was accepted to the IROS 2026 PWMS workshop. Meanwhile, the Franka dataset family surpassed 160K+ downloads on Hugging Face! Open Axis Benchmark Launch Built in collaboration with @openroboto, this living benchmark engine pulls from a library of 6,000+ tasks and 5.5M+ trajectories to accurately evaluate robot manipulation models without overfitting. Ecosystem and Campaign Growth The Binance Wallet campaign closed successfully with over 477K task completions and the Challenger Program was introduced for top-performing contributors. Just as LLMs were powered by the internet, the future of robot intelligence is being built on real-world data, on-chain on Base #AxisRobotics #Base #PhysicalAI #Robotics #Web3

凌晨三点还在盯盘的兄弟，举个手🙋 @sleepagotchi 我也是。手表戴了好几年，每天晚上都在记我的心率、深睡时长、半夜醒了几次，数据早就攒了一大堆。可这些东西现在躺在哪家公司的服务器里，值多少钱，我一概不知道，更别说分到一分钱。 这也是我最近又回头细看@sleepagotchi 的原因。 不少人对它的印象还停在"睡觉挖矿"，按时睡觉领 NFT、攒积分等空投。这一轮升级之后，它想做的事其实大了不止一圈：让你的睡眠和健康数据变成你自己能授权、也能分到收益的东西。官方叫permissioned data value，说成大白话就是，数据你说了算，别人想用得先经过你，好处也要有你一份。 有个细节我挺服气。它在Quests Hub 里用 SOL 钱包、邮箱、Telegram ID 把你以前在各个产品里留下的记录串到一起，Alpha 测试、V2 Beta 的连续打卡、Telegram 和 LINE 上的 Lite 版、NFT 活动都算数。所以不管你当初在哪个端玩，那段经历都不会白费，对一路陪过来的老用户来说，这算是给足了面子。 再往上一层，Sleep Coach、Wellness Coach、Meal Planner、Shopping Agent 这四个 AI agent 读的就是这份数据。昨晚睡得不好，它会帮你调整；身体缺点什么，购物这一步也能顺手接上。$CHI 夹在中间，管访问权限、奖励和结算。 我不敢打包票这条路一定走得通。健康数据太敏感，隐私和合规每一步都得抠得很细。但方向我是认的：Web2 一直把我们的身体数据当成免费原料，Web3 起码该让人先把所有权拿回来。 今晚早点睡吧，这回你睡出来的数据，总算有机会记在自己名下了 🦖 @sleepagotchi

The physical world is changing faster than maps can keep up. That’s why @vangrid_io is interesting to me. Instead of asking AI to rely on stale maps or simulated environments, Vangrid creates a path from real-world capture → 3D reconstruction → verifiable spatial data. And the demand side is getting more interesting: AI agents can now post capture bounties themselves, fund them in USDC through x402 on Base or Arc, and receive the resulting physical-world data after a contributor captures the requested location. No traditional account or API key is required for the agent flow. The contributor still gets a human role: walk the site, capture the environment, submit it, and get paid when the buyer accepts. The raw video isn't the end product. The reconstruction is. That distinction matters for Physical AI: a robot doesn't need a random phone video of a warehouse—it needs usable spatial information about the warehouse as it exists now. Vangrid is turning smartphones into a distributed perception layer while putting provenance and settlement onchain. The bigger loop is: AI requests reality → humans capture it → Vangrid reconstructs it → cryptography verifies it → USDC settles the work. Physical AI doesn't just need more intelligence. It needs fresh eyes on the ground. @vangrid_io is building them.

Metaplanet bán 10,000 $BTC rồi mua lại 11,000. Trong quý 3, công ty thu $790M từ bán 10,000 BTC và chi $950M mua 11,000 BTC, nâng tổng lên 44k BTC. Họ nói bán để chứng minh thanh khoản khi đang xin xếp hạng tín nhiệm, kèm chính sách mới: 85% đến 90% tài sản là BTC, vay không quá ~10% NAV. Theo số công bố, giá bán bình quân ~$79k còn giá mua lại ~$86k, nên vòng bán rồi mua có thể tốn vài chục triệu $ chênh lệch. Cơ quan xếp hạng cần thấy tài sản thế chấp bán được thật, không chỉ nằm trong ví. Trả vài chục triệu $ để lấy bằng chứng thanh khoản là phí vào cửa của hệ thống tín dụng truyền thống.