Brooklyn. |W

1.3K posts

Brooklyn. |W banner
Brooklyn. |W

Brooklyn. |W

@binlan883

金融的尽头是数据?或许是星象🌠 不空谈运势,只抓星象与金融的交叉点: •行星逆行期的市场避险指南 •星座能量对应的行业风口 •个人星盘里的投资优劣势 跟着星象做交易,让决策多一份宇宙视角📈 #玄学视角 #占星金融 Seismic Ecosystem Member | On my Magnitude Journ

Katılım Temmuz 2025
933 Takip Edilen410 Takipçiler
Brooklyn. |W
Brooklyn. |W@binlan883·
2026丙午年,劫在心火(眼睛也属火) 这一年,真正的养生, 不在于你跑了多少公里,不在于你练了多少块腹肌。 在于你守住了多少精血, 收住了多少心火,睡住了多少子午觉。 守恒,才是天道! 祝安好~
Brooklyn. |W tweet media
Brooklyn. |W@binlan883

今天金木三分相 让我们有了“刚好合适”,“刚好值得”的顺心和轻盈 土海合相白羊 理想落地,实干生财 敢想更敢干的人受欢迎 九紫离火运 不再是卷房子、车子、票子, 而是卷心态、卷睡眠、卷情绪稳定。 祝您前行如骏马,回望有温柔~

中文
0
0
0
34
Brooklyn. |W
Brooklyn. |W@binlan883·
📉 机器人数据新范式:从“刻意收集”到“劳动副产品” @NVIDIAGTC 现场洞察: ✅ 人类视频=预训练新石油 ✅ 2026=部署元年 ✅ 挑战不在模型,而在场景策略 别建数据农场,让机器人在工作中自我迭代。 这就是 @BitRobotNetwork 的数据飞轮。⚙️🧠
amanda@amanda0x

I spent the past week @NVIDIAGTC talking to robotics builders, investors, and researchers. And checking out 100+ robots around the venue. While the growing number of robotics data collection providers has become a meme, one thing is clear: robotics is still real-world data starved. 1) Learning from humans is now core to the data pyramid. In the "Humanoid Robots at Scale" panel, @chelseabfinn @aelluswamy @pathak2206 highlighted the value of human video for pre-training. Across teams, egocentric video and UMI/glove data came up repeatedly as key inputs. 2) Real-world deployment data is the gold standard. But, we’re early. 2026 will be the year of enterprise pilots and home robot betas. Jensen Huang said the ChatGPT moment for autonomous vehicles has arrived. This is after decades and billions of real-world miles. At GTC, the robot demos still break. 3) Deployment strategy is one of the hardest open questions. Homes → lower willingness to pay, higher safety constraints, more general-purpose systems. Commercial → higher willingness to pay, narrower tasks, often solvable with arms or quadrupeds before full humanoids. 4) The natural next step: robotics data companies facilitate labor. Over time, the most valuable data won’t be collected directly. It will be generated as a byproduct of real-world robotic labor, with models continuously evaluated in deployment. This is the path we’re building toward @BitRobotNetwork.

中文
2
0
8
76
kimi211🧸
kimi211🧸@AubrieHart17213·
收工了家人们,本来想进厂打螺丝给小美买个苹果17,结果上两天班 我发现原来我不是那么爱小美!
kimi211🧸 tweet media
中文
22
0
32
309
Brooklyn. |W
Brooklyn. |W@binlan883·
Who can say no to a tiny car that talks back? ✅ Sassy comebacks includedWe equipped the @frodobots $299 Mini+ Rover with an @openclaw AI brain: Full control over movement✅ Non-stop dad jokes Docs coming soon — get ready to build your own sassy little ride! @BitRobotNetwork
BitRobot Network@BitRobotNetwork

We built an @openclaw agent that drives the @frodobots $299 Mini+ Rover, cracks jokes, and talks back. Docs dropping soon so you can build your own

English
1
0
1
64
Meta
Meta@metanoetix·
With the large-scale development of AI, the explosive growth of AI agents is placing higher demands on blockchain infrastructure. When we talk about blockchains, words like “slow” and “expensive” often come to mind—especially on major networks like Ethereum and Solana. Improving transaction confirmation speed and data storage efficiency has long been a top challenge for technical teams. Optimum is a project designed specifically to address these issues. Rather than replacing existing blockchains (like Ethereum or Solana), it acts more like adding a dedicated GPU to a computer—enhancing performance through a “plug-in” approach that provides ultra-fast data transmission and massive memory capacity. Currently, Optimum consists of three core products, which can be understood as a “highway,” “super memory,” and “shared accelerator”: 1. Optimum P2P: The “Information Highway” for Blockchains This is Optimum’s flagship product, essentially a dedicated data channel built for validator nodes. Think of data packets as pieces of a large file. Traditionally, you must collect all original “puzzle pieces” to reconstruct the file. With RLNC (Random Linear Network Coding), the pieces you receive may look random, but as long as you gather enough of them—even if they’re not the original ones—you can reconstruct the complete file immediately. This dramatically improves transmission efficiency and reduces latency. Optimum emphasizes a non-intrusive integration strategy: instead of replacing existing clients, it operates as a sidecar, compatible with multiple execution environments such as EVM, SVM, and MoveVM. Validators can integrate it without changing hardware or upgrading their clients. It has already been tested on Ethereum’s Hoodi testnet, with 7 top-tier node operators integrating it—collectively representing about 15% of the network’s total staked ETH. 2. Optimum DeRAM: The “Super Memory” On-Chain If the first product solves data transmission speed, DeRAM addresses storage speed and capacity. Current blockchain data read/write performance is slow—like running a game on a mechanical hard drive. This makes it difficult for high-frequency trading, on-chain gaming, and AI inference—applications that require ultra-fast responses—to operate efficiently on-chain. DeRAM (Decentralized Random Access Memory) delivers a Redis-like experience (a high-speed in-memory database). It also leverages RLNC to ensure atomicity, consistency, and persistence of data operations. 3. Flexnode Network: A “Shared Accelerator” for Everyone Introduced in January 2026, this concept is highly disruptive. Today, blockchain acceleration is dominated by large node operators, leaving little room for ordinary users. Flexnode aims to change that by enabling individuals to contribute using personal devices—like laptops or idle desktops—to participate in a global data acceleration network. These devices help professional validators propagate block data faster. In return, contributors earn economic rewards. This effectively lowers the barrier to participating in “memory infrastructure” to near zero. By involving everyday users, the network not only becomes faster but also more decentralized—reducing reliance on a handful of large service providers. Only by overcoming fundamental challenges in transmission and storage can we unlock the true potential of the on-chain AI agent era. @get_optimum
Meta tweet media
English
1
0
2
32
Brooklyn. |W
Brooklyn. |W@binlan883·
快乐交给小龙虾,方向交给 Seismic。 Rocky 在线,Seismic 无限。 01 虾路相逢 勇者胜! @SeismicChina
Brooklyn. |W tweet media
中文
9
0
14
199
guava
guava@bale200099·
周末愉快😌
guava tweet media
日本語
9
0
21
217
Brooklyn. |W
Brooklyn. |W@binlan883·
DeFi Research Challenge: @pendle_fi’s points program pairs real yield from tokenized yield markets with lasting value Its fixed/variable yield AMM is core DeFi infrastructure, solving critical pain points. Own the future of DeFi, not just points @ORBT_protocol #DeFi
Brooklyn. |W tweet media
English
1
0
4
84
ORBT
ORBT@ORBT_Protocol·
Intent-based settlement is becoming the default execution layer of DeFi. Monthly intent volume: ~$30B Monthly DEX volume: ~$231B Intents are already 13% of the market and accelerating fast. ORBT is the unified liquidity layer powering intent-based settlement at scale.
English
211
179
499
21.2K
0xzll(❖,❖)
0xzll(❖,❖)@zllnevergiveup·
冰墙能挡下子弹,却挡不住我想和你一起在 Seismic 里赢下每一局的决心 @SeismicSys
0xzll(❖,❖) tweet media
中文
12
0
16
297
炎炎(✱,✱)
炎炎(✱,✱)@yanyanRialo·
时隔一年半 首骑 是不是也要感谢web3 哈哈 2026锻炼身体 好好赚钱 一路长虹
中文
14
0
27
614