这一波牛市，怎样寻找赚钱机会？

深潮Pubblicato 2025-08-11Pubblicato ultima volta 2025-08-12

OnChain+OffChain，TradiFi 成主线大叙事。

撰文：Haotian

都喊着牛来了，但可否知道这一波寻找市场 Alpha 和 Beta 的方法论完全不一样了，谈几点观察：

1）OnChain+OffChain TradiFi 成主线大叙事：

稳定币基础设施化：稳定币成连接传统金融与 DeFi 基础设施的「血液」，锁定稳定币的跨链流动，APY Yield 差异以及新型创新拓展；

BTC/ETH 微策略「币股」效应：上市公司将加密资产纳入资产负债表成为趋势，寻找具备「准储备资产」的潜力优质标的成为关键；

To 华尔街创新赛道崛起：专为机构设计的 DeFi 协议、合规收益产品、链上资管工具将获得巨额资金青睐。原来的「Code is Law」成了全新的 「Compliance is King」；

2）Crypto 纯原生叙事去伪存真加速：

以太坊生态迎来一波复兴潮： $ETH 价格突破会重燃以太坊生态的技术叙事的创新潮，取代过去的 Rollup-Centric 大战略，全新 ZK-Centric 会成为主旋律；

高性能 Layer1 实力比拼：不再是 TPS 竞赛，而要看谁能吸引真实的经济活动，核心指标包括：稳定币 TVL 占比、原生 APY 收益率、机构合作深度等等；

山寨币最后的黄昏：普遍性的大山寨季存在资金量动能不足的根本难题，局部山寨币会出现「死猫跳」行情，这类标的特征：筹码集中度、社区活跃度、能否搭上 AI/RWA 等新概念；

3）MEME 币从投机工具升级为市场标配：

资金效率：传统山寨币市值虚高、流动性枯竭，MEME 币以其公平发射、高换手率成为资金新宠，将抢占大部分濒死山寨币市场份额；

注意力经济主导：KOL 影响力、社区文化建设、热点 FOMO 模式仍是核心竞争力，流动性分配依然遵循注意力法则；

公链实力新指标：MEME 币市场活跃度会是衡量公链综合实力的重要标准。

Letture associate

Robotic GPT-3 Moment Shakes Silicon Valley: Zero Lines of Code, Learns Instantly, with Investments from Jensen Huang and Fei-Fei Li

Generalist AI's newly released robot foundation model GEN-1.5 is being hailed as the "GPT-3 moment" for embodied AI. The model demonstrates remarkable one-shot and few-shot learning capabilities in physical manipulation tasks. By watching a single 3-12 second demonstration video (physical prompting) without any training or code, the robot can attempt the task with a 59% average success rate across 10 tasks. With just 5 minutes of demonstration data and minimal fine-tuning (10 gradient steps), the success rate jumps to 83%. The most significant breakthrough is the emergence of spontaneous, improvisational problem-solving abilities not present in the training data. For instance, after learning to sweep blocks with a brush, the robot can adapt to use a banana similarly or switch to a completely new "scoop-and-pour" strategy when given a dustpan. Other emergent behaviors include removing obstacles, correcting errors, and spontaneously sorting objects. These capabilities stem from over 8 months of large-scale pre-training on physical interaction data, suggesting the existence of a Scaling Law for embodied intelligence—where model generalization improves with more data and training time. This approach drastically reduces the cost and expertise needed to teach robots new skills, potentially democratizing robot programming. The release coincides with a surge in the humanoid robotics sector, marked by events like the World Robot Conference and significant investments. While the demonstrated tasks are still relatively simple, the scaling trend and emergent behaviors point toward a future where general-purpose robot "brains" could be easily adapted to various hardware "bodies," reshaping the industry's competitive landscape.

marsbit7 min fa

Robotic GPT-3 Moment Shakes Silicon Valley: Zero Lines of Code, Learns Instantly, with Investments from Jensen Huang and Fei-Fei Li

marsbit7 min fa

Semiconductor Industry Sees a New Wave of Price Hikes, with STMicroelectronics, Maxscend and Others Increasing Prices Intensively

The semiconductor industry is experiencing a new wave of price increases driven by supply-demand dynamics and rising costs. Throughout August, multiple international and domestic semiconductor companies have issued price adjustment notices to clients, with a focus on analog and RF chips. Key companies adjusting prices include STMicroelectronics, which is implementing its third price hike in 2026 effective August 23, citing sustained demand pressure and rising costs across the supply chain. Analog Devices, Inc. (ADI) has also announced a second price increase for 2026, effective September 13, due to "unprecedented" demand growth and escalating manufacturing costs. In China, RF chip leader Zhuosheng Microelectronics issued a price adjustment notice for its full RF product line, effective September 1, its first hike this year, attributing it to rising raw material and foundry costs. Additionally, Chinese MCU manufacturer Nations Technologies announced a 10-20% price increase for some products from October 1. This new round of price hikes has shifted its focus to mature process nodes (e.g., analog, RF, power, MCUs), partly due to AI computing demand consuming 8-inch wafer capacity. This contrasts with the first half of 2026, when the surge was led by AI memory chips like HBM and server DRAM. Analysts note that the long lead time for expanding analog/power semiconductor capacity (18-24 months) may sustain price resilience in the short term. However, downstream manufacturers' cost tolerance and fluctuating global demand could become constraints, potentially limiting further significant price increases. Companies have linked the hikes to ensuring future supply stability by funding capacity expansion.

marsbit10 min fa

Semiconductor Industry Sees a New Wave of Price Hikes, with STMicroelectronics, Maxscend and Others Increasing Prices Intensively

marsbit10 min fa

Cursor Launches the Strongest "Automatic Workhorse" Mode, /goal Dominates the Scene

Cursor's latest major update supercharges its Cloud Agents with autonomous, persistent workflows. The headline feature is `/goal`, transforming one-off tasks into continuous objectives—like fixing all flaky tests until CI stays green—where the Agent works unattended until the goal is met. New "Subscriptions" allow cloud-based Agents to autonomously monitor event sources like GitHub PRs, Slack threads, or schedules. An Agent can now own a PR from creation through merge, automatically handling CI failures or bot comments without human intervention. The update also enables sub-agents to run in isolated cloud VMs with their own project copies, allowing parallel work without conflicts. Additional tweaks include "Custom Modes" to lock an Agent into a specialized skill and "Steering," which queues user interruptions until the Agent's next tool call. Cursor reveals that over 35% of its internally merged PRs are now autonomously created by cloud Agents, up from 30% in February 2026. Accompanying this shift, the "Auto" pricing tier will transition to model-based billing starting August 24, making the cost of running these high-autonomy agents more granular. This evolution moves the developer's role from writing code to managing AI agents that handle coding, PRs, CI, and testing—fundamentally reshaping the software development pipeline.

marsbit10 min fa

Cursor Launches the Strongest "Automatic Workhorse" Mode, /goal Dominates the Scene

marsbit10 min fa

MemeCore to List on Nasdaq as Part of $1 Billion Deal

MemeCore has entered into a $1 billion agreement with publicly-traded company ZeroStack (NASDAQ: ZSTK), which develops decentralized AI infrastructure. Key ecosystem participants Puple AI Inc. and Blockcat Pte. Ltd. are involved in the deal. As part of the agreement, these companies will transfer 925.9 million MemeCore ($M) tokens to ZeroStack at $1.08 per token, totaling $1 billion in assets. In return, they will receive 3.5 million shares of ZeroStack and warrants to purchase up to an additional 36.2 million shares. The warrant exercise price is set at $25.19, over 12 times ZeroStack's recent market price. Issuing the additional shares requires shareholder approval per Nasdaq rules and will be subject to a lock-up period of up to 10 years. This transaction will allow MemeCore to go public and link its meme culture-based blockchain ecosystem with ZeroStack's decentralized AI focus. The companies also plan to explore collaboration in decentralized AI, digital assets, consumer applications, and blockchain economics.

cryptonews.ru12 min fa

MemeCore to List on Nasdaq as Part of $1 Billion Deal

cryptonews.ru12 min fa

Solana cuts blockchain slot time to 350 milliseconds

Solana has reduced its blockchain slot time to 350 milliseconds, the first reduction since the network's launch. According to Solana Foundation VP of Technology Jacob Creech, this marks a "new era" with a next target of 300ms. Average slot times are currently 360ms, down from the original 400ms target. The reduction is part of a broader plan to eventually cut slot times to 200ms to improve network latency and confirmation speed. This initial cut is the first of four planned 50ms reductions. The changes, formalized in the approved SIMD-0525 proposal, are slated for implementation in the upcoming Agave v4.2 validator client, though the schedule remains tentative.

cointelegraph12 min fa

Solana cuts blockchain slot time to 350 milliseconds

cointelegraph12 min fa

Trading

Spot
活动图片