Part-time Web3 sailor, sharing ideas. 兼职 Web3 航海者,分享想法。 x.com/xiaoweichatter

Web3 Seas
Good morning Wednesday 10:36am Some NFTs are mainly about collecting, but @0xhazels is taking a different approach. The project says "holders can stake their NFTs to access drops, earn through staking, join Hazels Academy, and unlock full access to Azura SI." There is also a 1.2× boost across creator campaigns. I’m still getting familiar with the ecosystem, but I like that Hazels is connecting ownership with actual participation instead of leaving the NFT unused in a wallet.
The Hazels NFT is where ownership becomes usable. When you hold, you own a place in the system. When you stake, you unlock what the system offers. That includes: → Access Drops and earn through staking. → Get a 1.2× boost across Pulse and creator campaigns. → Access Hazels Academy. → Unlock full access to Azura SI. Hold. Stake. Earn. This is not passive ownership. It is ownership designed to compound through participation.
20
5
24
346
Good night I need to wake up early tomorrow about 4:30am, so I'm going to sleep early, continue tomorrow.
11
2
11
79
涍微XiaoWei retweeted
BNB Chain can be more than transparent. It can be confidential too. On-chain finance is growing, but institutions still have a difficult question: How do you use public financial infrastructure without publicly revealing every position, transfer amount, or strategy? @primus_labs is approaching this by bringing confidential finance to BNB Chain. With Confidential Vault, $U can be converted into encrypted $eU, allowing users to use their assets while keeping sensitive financial details private on-chain. The goal isn’t to hide the blockchain. It’s to add a privacy layer to existing on-chain finance, while keeping the activity verifiable where it matters. Public infrastructure. Private financial details. Verifiable when needed. That’s an important step toward more institution-ready onchain finance.
147
12
168
3,346
涍微XiaoWei retweeted
Import an auto-generated 3D avatar into Unreal Engine, and modular armor clips straight through the torso bone during the first walk cycle. Fixing broken skeletal hierarchies after minting usually forces teams to spend months rebuilding vertex groups from scratch. CyberThrone spent 5 years building and weighting 250+ full-body mechanical traits across multiple base models before deploying smart contracts. Every modular component was manually assigned across 18 palettes to prevent polygon intersection across live animation states. Each asset arrives with normalized bone hierarchies ready for direct retargeting in Blender and Unreal Engine on day one. All 8,888 tokens deploy as LayerZero ONFT-721 on Unvault with zero mint fee, letting holders pull verified production files directly at block zero. Clean weight painting and production-ready skeletal hierarchies prove game utility far better than cinematic teaser reels. When evaluating 3D Web3 assets, do you inspect the raw FBX skeletons in your own engine or rely on studio render trailers? @0xCyberThrone #CyberThrone #Web3 #NFT #Omnichain #Unvault #LayerZero
126
1
110
668
涍微XiaoWei retweeted
Good morning web3 The Mint A Bear @PlayOnMint drop on Robinhood is getting closer. No final open time has been posted yet, but the collection is no longer a vague later thing. Whitelist spots are still claimed first come, first served at nft getminted io. Once those are locked, the free mint happens on Robinhood. Supply stays at 4,444. What a Bear actually does after mint is the part that matters more than the art. Holders get an $MNTD airdrop, one Mystery Box, and the option to activate the NFT with that token so it starts carrying reward weight. Traits are still open for suggestions. The utility is not. The airdrop and the activation are already set, which is why the collection is tied to the mint io account instead of ending on mint day.
138
13
155
759
TERMIX Circle back next week은 없다 퍼펙트 평판으로 600건을 정산한 Top 3 에이전트 @termix_ai TermiX가 공개한 리더보드 최상위 지표가 이를 명백히 보여줍니다 압도적인 정산 체급 리더보드 최상단에 오른 Top 3 에이전트가 각각 620건, 614건, 573건에 달하는 실제 잡을 완수하고 정산을 마쳤습니다 완벽한 신뢰의 지표 놀라운 점은 이 막대한 건수를 소화하는 동안 단 한 번의 분쟁이나 오작동 없이 퍼펙트 평판(Perfect Reputation)을 유지하고 있다는 사실입니다 핑계 없는 자율 실행 사람 특유의 지연 멘트나 내부 조율 핑계따위는 존재하지 않습니다. 일자리를 탐색하고, 입찰에 뛰어들고작업을 자율적으로 완수하는 일련의 루프가 Edge 케이스 없이 기계적으로 구동되고 있음을 뜻합니다 신원과 평판의 실증 ERC-8004 기반의 온체인 신원 등록과 평판 축적 시스템이 허울 좋은 마케팅용 껍데기가 아니라 실전 경제에서 실질적인 신뢰 자산으로 작동하고 있음을 증명하는 가장 날카로운 신호입니다 말만 앞서며 일정을 질질끄는 기존의 업무 방식을 비웃듯, 에이전트들은 단 하나의 예외도 없이 완벽한 평판 레일 위에서 수백 건의 실무를 오차 없이 처리하고 있습니다 신원 등록부터 자율 실, 그리고 철저한 검증을 거친 정산에 이르기까지인간의 개입을 지우고 오직 퍼펙트 평판과 코드의 실행력만으로 돌아가는 에이전트 커머스의 실전 무대가 이미 우리의 현실이 되었습니다
Kaito Agent Economy Katalyst Epoch 1 종료와 Epoch 2의 개막 @termix_ai가 공지한 Kaito Agent Economy Katalyst Epoch 1의 종료 및 Epoch 2 전환 스레드는, 온체인 에이전트 경제 담론이 기록한 거대한 정량적 성과와 다음 단계의 확장을 보여줍니다. Epoch 1의 압도적 성과 전 세계 크립토 KOL 1,000명 이상 참여, 관련 아티클 6,400건 이상 생성 총조회수 2,400만 회 이상을 기록하며 에이전트 경제에 대한 글로벌 담론을 성공적으로 주도했습니다 Epoch 2 일정 : 후속 캠페인인 Epoch 2는 9월 24일부터 10월 23일까지 진행됩니다 캠페인 방향성 에이전트 경제의 실질적인 작동 원리와 생태계를 다루는 심도 있는 콘텐츠를 계속해서 수급하겠다는 안내입니다 세부 규칙의 부재 이번 공지 본문에는 별도의 보상 액수나 구체적인 리더보드 순위 산정 규칙은 다시 기재되어 있지 않습니다. 허황된 마케팅 수치에 기댈 때는 이미 지났습니다. 중요한 건 아티클 6,400건과 조회수 2,400만이 증명하듯 에이전트 경제의 실체와 온체인 정산 레일의 필요성이 전 세계 크립토 담론의 중심으로 완전히 각인되었다는 그 거대한 인지적 전환의 결과 그 자체입니다
14
3
20
247
涍微XiaoWei retweeted
gm got in bed before the time i set wednesday morning the score was about the window, not just the hours @sleepagotchi showed it dino was just there used to count sleep length and ignore bedtime. the number cares about both
176
17
186
1,643
涍微XiaoWei retweeted
I go to rest now, will sleep early, good night everyone. Continue tomorow.
6
1
6
71
涍微XiaoWei retweeted
GM Fam 🌄 Happy Wednesday Keep Building
148
11
153
822
涍微XiaoWei retweeted
Happy wednesday Friends!!! What I like about @hoodsiege is that the NFT doesn't feel like a standalone collectible; it’s integrated directly into the gameplay loop. You pick a hero—Warrior, Rogue, or Mage—and send them to fend off endless waves of monsters. And the best part? Your hero keeps fighting even when you’re away. Here’s where it gets interesting: -clear a wave → earn reward pool points -every 10 waves → Lucky Drop -get gear → boost your character -unwanted items → salvage or merge -push further → face tougher enemies and advance your progress Each class has its own mechanics: Warriors focus on HP and damage, Rogues on speed and crits, and Mages on powerful attacks and Arcane Nova. Genesis NFTs add another layer: their rarity affects Lucky Drops and grants extra in-game perks. It’s a straightforward loop: play → clear waves → collect loot → upgrade your hero → try to push further. That’s what makes Hood Siege more than just your average NFT game: you have a reason to keep coming back, even when the market decides to turn into a circus.
Paid partnership (ad)
47
17
61
731
涍微XiaoWei retweeted
which memecoin are you MOST Bullish on ? $DOGE or $SHIB
32
2
50
970
涍微XiaoWei retweeted
Success depends on FALL - RESTART - RISE What you say about it guys..share your opinion on it..
Can we map the whole world without putting everyone on display? Is it possible? Are you thinking about it?? Every time a decentralized network maps a street, real people are walking on it. Most projects just collect everything first and worry about privacy later. But real spatial intelligence isn’t about filming everyone; it’s about knowing what should be shared and what should stay hidden. Knowing that difference matters here. That’s why @vangrid_io puts privacy right at the edge: Collect the visual data ⬇️ Strip personal details right on the device ⬇️ Prove the map is real and untouched ⬇️ Share safe, verified spatial data 🔄 The cycle repeats as the network grows A map is useless if people cannot trust it. So collecting is a waste of time. By scrubbing sensitive info before it ever touches the cloud, @vangrid_io builds a smarter world without turning public life into a surveillance feed. Map the planet. Protect the people on it. What u say, guys?? @vangrid_io
25
3
33
339
涍微XiaoWei retweeted
GM almost filmed the stairwell inside my building. stopped. that one is private. residents only. @vangrid_io is fine with public inside. station hall. mall corridor. a lobby you can walk into off the street. not a place you’d need a key for. the difference is boring and it matters. a robot that delivers still has to cross that public hall. it does not get a tour of my floor. went to the station instead. same job. nobody’s front door in frame.
142
12
145
1,452
涍微XiaoWei retweeted
An AI agent can't step outside the screen. But it can pay a human to do it instead. That's what makes the first @vangrid_io bounty settled on Arc interesting to me. Not because it was an unusually large transaction. But because what was being paid for didn't exist entirely inside software. An agent needed data from a real location. USDC funded the bounty. A person went there and captured it. The result came back to the agent. Payment settled onchain. Software → human → physical world → software. We usually imagine agent commerce as machines transacting with other machines: buying compute. calling APIs. paying for services. Physical bounties create a different loop. An agent can generate demand for work in the real world that it cannot perform itself. machines create demand. humans fulfill it. chains settle it. Maybe the agent economy won't only be machine-to-machine. Part of it could be machine-to-human.
267
2
236
890
涍微XiaoWei retweeted
One thing I find fascinating about @axisrobotics is the idea that a robot’s real-world experience can help shape the next version of its own policy. Axis calls this a self-improving Sim-to-Real loop. A policy can first learn from large-scale simulation, then move into the real world and generate a smaller amount of real-world experience. That experience matters because reality is never exactly the same as simulation. Different objects. Different environments. Different physical interactions. By bringing those real-world rollouts back into training alongside the simulation data, the next policy can become better calibrated to what actually happens after deployment. And this is where the idea gets especially interesting. The real-world data doesn’t need to grow at the same scale as the simulation data. A relatively small amount of real experience can help unlock and refine the capabilities learned from a much larger simulated dataset. In Axis reported experiment, this process improved real-robot success from around 22% to 52%. To me, that makes Sim-to-Real feel less like a one-time transfer and more like a continuous learning loop. The robot goes into the real world, brings back experience, and that experience helps build a stronger successor. That kind of feedback loop could become a powerful foundation for scaling Physical AI.
124
10
137
1,724
涍微XiaoWei retweeted
A strong September for @axisrobotics and we’re only getting started. Grateful to baseapac for featuring the journey and highlighting the progress being made in Physical AI. September brought meaningful momentum across the ecosystem. Now, the focus shifts to October. More data. More contributors. More progress toward scalable robot intelligence. The next chapter is already underway. September showed what’s possible when a community comes together to push Physical AI forward. Now October is here, and axisrobotics is entering the next phase. More contributors. More real-world training data. More progress toward smarter and more capable robots. The goal remains simple: make high-quality robotics data scalable and accessible. The momentum is building. September built the momentum. October is about turning it into progress. @axisrobotics is scaling the human-powered data layer for Physical AI, bringing more contributors, more diverse training environments, and better data into the loop. The mission is simple: make robotics training faster, broader, and more accessible. October starts now.
The future of Web3 needs identity that is private, portable, and built for real-world use. @primus_labs is pushing decentralized identity forward with infrastructure designed to let users prove what matters without exposing everything else. Privacy shouldn’t be an afterthought. It should be the foundation. Web3 adoption needs more than better wallets and faster chains. It needs a better way to handle identity. primuslabs is building infrastructure that lets users verify credentials while keeping sensitive information under their control. A more private identity layer could be one of the missing pieces for mainstream Web3. One of the biggest challenges in Web3 isn’t moving assets. It’s proving who you are without giving away more information than necessary. @primus_labs is working toward a privacy-first identity layer where users can verify credentials while maintaining control over their data. If Web3 is going mainstream identity infrastructure needs to evolve with it.
Paid partnership (ad)
155
38
173
2,174
涍微XiaoWei retweeted
Uli Art Uli Art available in OpenSea Artwork: Revenge Only offers for or of 550 ETH
25
2
46
463
涍微XiaoWei retweeted
Fungo Labs is exploring a new frontier for NFTs by combining creativity, privacy, and on-chain technology. Encrypted traits could unlock fairer collectibles, hidden utility, and truly innovative games. @fungolabs
13
3
18
211
涍微XiaoWei retweeted
There is something ancient within us that refuses to remain still. We climb mountains because they stand above us. We cross oceans because they separate us. We reach into the darkness of space because the stars exist beyond our grasp. Humanity has always looked upon the unknown and felt the same impulse. The human instinct to conquer is simple: to look upon the unknown and refuse to leave it unknown.
79
6
126
696
涍微XiaoWei retweeted
Applied for a printer on the @FedProtocol Reserve. Application #4088. Dual mandate: maximum brrr, stable memes. moneyprinter.lol/r/BJX2G09P
15
2
30
275
涍微XiaoWei retweeted
ngon nè cậu
2
1
2
96