智能体 · 642

Tencent's chip leader Gao Jianlin left his job and started a business to target the RISC-V high-performance AI CPU circuit

According to MaxForAI, according to MaxForAI, the core head of Tencent's chip research and development, has recently left Tencent and started a business. He plans to develop high-performance CPUs based on the RISC-V architecture around high-performance AI servers and agentic AI (intelligent AI). Gao Jianlin is regarded as one of the early core promoters of Tencent's self-developed chip system. According to data, he formed an FPGA hardware team within Tencent in 2013, began setting up AI chip research and development in 2018, established the Penglai Laboratory in 2020, and promoted Tencent's development of various AI chips and data center deployment. This startup focuses on CPUs rather than the currently competitive AI GPU market. According to the report, Gao Jianlin believes that with the rapid development of Agentic AI, the AI inference process will involve model call, tool execution, search, database interaction, and large-scale task scheduling, and the CPU will assume a more important scheduling and control role in the AI system. According to reports, Gao Jianlin was involved in RISC-V related research and development during his time at Tencent, and participated in various technical directions such as chip architecture, verification, and back-end. Its new company plans to build high-performance server CPUs based on the open instruction set RISC-V to enter the AI infrastructure market. Currently, the name of Gao Jianlin's new company, financing conditions, and specific product launch dates have not been disclosed. The market is concerned about whether it will become another emerging force in the field of AI chips in China targeting server CPUs and smart body infrastructure.

13h ago

They are all stealing earlier data. Where exactly is VC Alpha hidden?

Author: insights4vc Compilation: Shenchao TechFlow Original title: Private Equity Market Intelligence Warfare Heats Up: In the AI Era, Where Did VC Alpha Come From? Guide to Deep Wave: Venture capital returns are extremely concentrated, and finding a good company in the early stages is almost the life and death line of a fund. This article breaks down the latest evolution of private equity market data tools and whether they can actually bring in excess profits. This is a sobering map for investors who are using AI and research tools to find projects. Venture capital has always been an information business. The advantage often lies in timing: founders tell former colleagues instead of updating data first; new companies start recruiting people before they appear in the database; investors start watching a team before the funding is announced. This advantage is important because VC returns are highly concentrated. According to data from the 2026 Oxford Academic Study, 4.5% of the investment amount contributed to a return of about 60% in a long-term LP data set. [1] Therefore, missing a few excellent companies can affect the entire fund. But finding them early is only part of the problem. Investors also need to develop beliefs, get credits, obtain meaningful holdings, and keep things right for a few years. The private equity market data industry is now getting closer to the moment the company was born. PitchBook, Crunchbase, Dealroom, Tracxn, and CB Insights remain core recording systems for transactions, funds, valuations, and company history. PitchBook generated revenue of $174.7 million in the second quarter of 2026, equivalent to nearly $700 million in annualized revenue. [2] The new platform is not replacing this layer. They're extending this layer with faster updates, behavioral data, and signals that predate traditional company records. Three changes stand out the most. First, companies such as Harmonic and Specter are building a continuously updated map of companies and people, rather than relying mainly on regularly updated data. Second, specialty products are looking for earlier behavioral signals. Evertrace tracks metrics formed by founders, including company registrations, technical activity, research, and domain names. Frontrun monitors changes in selected venture capitals' interest maps on X. Third, the API and Model Context Protocol (MCP) are moving this data into the fund's own software and AI workflows. Crustdata represents the infrastructure side of this market, while Affinity complements first-party relationship data from emails, calendars, and CRM events. Adoption is visible, but evidence of excess return on investment is not clear. Harmonic says hundreds of venture capital teams use its platform, and Specter reports more than 300 investment institutions, Evertrace more than 200 funds, and Affinity more than 3,300 private equity firms. Listed company Tracxn disclosed that it had 2,289 customer accounts in fiscal year 2026. [3] [4] [5] [6] Most of these figures are self-reported by companies. Vendors rarely disclose the complete set of companies unearthed by their models, making it difficult to assess accuracy, recall rates, false positives, and the economic value of individual leads. No single signal alone is enough. Employee departures may be early but vague. Company registration is objective but common. GitHub activities are valuable in developer-led markets, but have limited relevance in other areas. Hiring speed and employee migration provide broader signals, while revenue, customer, and usage data are often more valuable for decision-making, but come later. When several credible industry experts focus on the same company, investors' attention can provide early signs, even though this signal is platform-dependent and may reinforce itself. The strongest defensive sources are likely to be hidden deeper in the data stack: historical time series that cannot be reconstructed later, accurate physical analysis across people and companies, authorized first-party fund data, and distribution through CRM systems, APIs, and agents. Public data is not necessarily proprietary. However, five years of correctly time-stamped change history can become a proprietary asset. AI is more likely to make these infrastructures more easily queried rather than eliminate the need for them. As research, classification, and workflow costs drop, clean data, sources, and institutional context become more valuable. Investment decisions, quotas, and relationships are still not something a simple layer of automation can solve. The likely outcome is that a broader market for private market intelligence will emerge, rather than an independent search for project software categories. A mature database will increase discoveries and...

1d agoburnking

Ethereum Foundation Launches Better.Codes Challenge to Advance Hashing SNARK to Prove Safe

Comparatively, better.codes, an open automated research challenge created by the Ethereum Foundation's formal verification team in collaboration with Yukon and zkSecurity, is now live. The platform formalized the self-contained problem in the Proximity Prize into Lean and placed KoalAirS12's machine inspection reliability community on a public leaderboard for anyone to drive improvements to advance hash-based SNARK and post-quantum Ethereum-related security benchmarks. Solvers can bring their own AI agents to prove a higher reliability lower bound for this Reed-Solomon proximity problem and move towards a fixed 128-bit goal. Lean Kernel verifies each submission, and the promoted certification enhances the open community, and its new theories, proof techniques, and unlikely results are synchronized upstream for reuse by all participants. Most hashed SNARKs in production environments rely on relevant proximity gaps and related agreed conclusions, and currently the verifiable results are still below the benchmarks the researchers believe in. This challenge aims to close this gap in an open, incremental, and verifiable manner. KoalAirs12 is derived from a related paper and formalized end-to-end in ArkLib. Participants can log in and clone the challenge repository via GitHub and submit it under a fixed theorem statement and verification framework; after confirmation by the comparator and Lean kernel, the results are recorded in the public repository and the solver and model used are indicated. Launched today is a reliability challenge that raises the KoalAirS12 certification lower bound to 128 digits. More topics may be added later. The rules are subject to the project terms.

1d ago

Privacy blockchain project Beldex raises $8 million, led by Sigma Capital

In comparison, according to Decrypt, the privacy blockchain project Beldex announced the completion of $8 million in financing, led by Sigma Capital, and institutions such as NTC, Nxgen, Digital Consensus Fund, and EAK Ventures. The new funding will be used for developer tools, privacy applications, protocol security, AI infrastructure construction, and ecosystem expansion. Beldex is exploring the use of its privacy technology to protect the identity, payment, communication, and data processing processes of AI agents, including research on technologies such as privatizing AI identities, encrypted AI communications, confidential payments, secure AI execution, and fully homomorphic encryption (FHE) and secure memory.

2d ago#financing

Alibaba's Wu Yongming: Ali AI's annualized revenue exceeds 49.5 billion yuan

Comparing news, Alibaba announced that Alibaba Cloud is undergoing a full upgrade to an intelligent cloud. In the next few quarters, AI and cloud business revenue growth will further accelerate. Alibaba Group released financial results for the first quarter of the 2027 fiscal year. During the analysts' conference call, Group CEO Wu Yongming said that AI has become the core engine for Alibaba Cloud's accelerated growth. This quarter, Ali's AI-related product annualized revenue (ARR) surpassed 49.5 billion yuan ($7.3 billion), and its share of Alibaba Cloud's external commercial revenue rose to 35%. The gross margin of AI-related products is significantly higher than the average for cloud products.

2d ago

Deagentai's new official website was officially launched, and the AI Agent hosting platform and the first application Sentry were released simultaneously

Comparing news, the decentralized AI infrastructure Deagentai announced the official launch of its new official website. The AI Agent hosting platform, which was launched simultaneously with the official website, is committed to providing long-term underlying support for the construction, deployment and permanent hosting of agents. As the first exemplary application launched on the platform, Sentry has been launched simultaneously to enable 24/7 permanent strategy monitoring, covering all Hyperliquid 308 targets (including crypto assets, stocks and commodities), and relies on the underlying trusted infrastructure to transform market insights into objective, automated strategy execution without emotional interference. Deagentai said that Sentry is only the first step in implementing the platform, and AI agents with more vertical scenarios will continue to be built and hosted on this platform in the future. At the same time, with the full implementation of the AI Agent platform, Sentry, and enterprise solutions, Deagentai has officially launched a $AIA programmatic normalized repurchase and destruction mechanism based on real agreement earnings to promote token deflation and value capture through actual business growth, and continue to feed back the long-term value of $AIA.

2d ago

Yao Shunyu reorganizes Tencent's multi-modal route: closer to Liang Wenfeng and away from Li Feifei

Text | Miao Zheng Editor | Wang Jing Source | Letter AI Tencent's mixed element multi-modal team has undergone another personnel change. According to media reports, Lin Xudong, who was responsible for xAI's multi-modal understanding, has left xAI and joined Tencent's mixed element as the head of the multi-modal content generation algorithm. The reason this personnel news is worth paying attention to is that it takes place in the context of continuous adjustments of mixed and multi-modal teams. Over the past period of time, news of the departure of the person in charge, the transfer of researchers, and the addition of new members came out one after another within the mixed yuan. Hu Han, the former head of multimodal understanding, left his career to start a business, and Tian Yonglong and others joined Tencent. The reporting relationship between the original multi-modal team also changed with the integration of the big language model department and the multimodal model department. However, does this mean that Tencent's multi-modal team is “changing the dynasty” is currently unable to draw a direct conclusion. What can be confirmed by public information is that mixed forces have indeed experienced personnel movements and organizational restructuring. The rumor of Lin Xudong's addition is more like a new signal in this adjustment: Tencent is recombining the two routes of multimodal understanding and content generation. So the question is, what exactly did Lin Xudong come from, and what abilities can he add to Tencent? And is Tencent's multi-modal approach shifting from “generating content” to Yao Shunyu's more biased “understanding context and acting in the world”? What is Lin Xudong's origin and what can he do after joining Tencent? According to public information, Lin Xudong graduated from Tsinghua University in 2018 and then went to Columbia University to study for his doctorate. While studying at the blog, his research interests included embedded learning, video analysis, and generative models. He also participated in the Vx2Text project in collaboration with Columbia University and Facebook AI. V indicates video, x indicates unknown, can be sound, voice, or even ambient sound. 2 represents TO, and Text represents subtitles. Its logic is to first convert different modes such as video and sound into vectors similar to “language tokens”, then uniformly feed the language model for fusion, and finally generate open text by an autoregressive decoder. Transformer can only understand tokens, so AI essentially doesn't understand video and audio file formats, making it even less likely to convert them into text. For example, if a dog jumps into the water next to a swimming pool, Vx2Text's video recognizer (V) will output keywords: dog, jump, pool; sound reader (x) will output: sound of water, fluttering. Although the product function of Vx2Text is “generation,” the core difficulty of the product is “understanding.” Of course, Vx2Text doesn't simply “translate” a screen into a few sentences. Models need to recognize people, objects, movements, and events from videos, understand how these things change over time, and finally organize visual information into language. After graduating from his PhD, Lin Xudong joined DeepMind and participated in Gemini-related multi-modal pre-training and post-training work. In 2025, he also joined xAI. According to public information, it is responsible for the direction of multimodal understanding and participating in the training of multimodal content understanding and generation models. Now that he has joined Tencent Hybrid, he will be responsible for the hybrid multi-modal content generation algorithm. Lin Xudong was added not so much to improve the performance of mixed-element multi-modal generation, but rather to solve a problem that plagues all multimodals — understanding. The previous generation model was more like a picture maker. Give it a hint, and it can generate an image or a video. But as long as users make more complex requests, the model just can't keep up. For example, the characters change in the long video, the shape of the object is not consistent before and after, the camera movement does not match the spatial relationship, etc. It's not because the model doesn't generate, but because it doesn't remember and understand the world steadily. Therefore, putting Lin Xudong in the position of multi-modal content generation is probably because he “translated” multi-modality into something AI can understand. Lin Xudong's addition can only be clearly seen in a larger context. That is, now Tencent's mixed element is reorganizing its multi-modal route. In January 2025, Tencent Outstanding Scientist (Tencent Distinguished Scientist) Hu Han succeeded Liu Wei, who had previously left his job, and was fully responsible for the research and development of mixed-element multi-modal models, and also served as Tencent's mixed-element big model Tech Lead. Tencent's internal organization was adjusted in the second half of 2025. He transferred from the Multimodal Model Department to the “Frontier” Frontier Technology Research Group under the Big Language Model Department. The title was changed to Head of the Multimodal Understanding Direction, and the reporting line was also changed to report to Yao Shunyu. The actual position changed from “the head of an independent department” to a “big language model...

2d ago字母AI#AI #Li Feifei #Liang Wenfeng #Tencent

Bitwise CIO: AI agents and tokenizing assets may drive a 50-100x increase in blockchain transactions

Comparing news, Bitwise Chief Investment Officer (CIO) Matt Hougan said that the market may seriously underestimate the scale of future blockchain trading activity, and as real-world asset tokenization (RWA) and artificial intelligence agents (AI Agents) reshape the financial market, on-chain transaction volume may increase 50 or even 100 times in the future. In his latest investment memorandum, Hougan pointed out that current crypto investors have three major misconceptions, one of which is underestimating the future coverage of blockchain applications. He believes that traditional assets on the chain and the participation of AI agents in transactions will significantly increase the scale of on-chain activity, and the tokenized stock market alone may lead to a 10-fold increase in trading volume. He pointed out that at present, the traditional stock market is usually only open from 9:30 to 16:00 EST on weekdays, with a weekly trading time of about 33 hours; while tokenized stocks can be traded around the clock, increasing the available trading time to 168 hours per week. If AI agents automatically execute transactions on behalf of investors in the future, the frequency of transactions may increase further. However, Hougan also acknowledged that the increase in transaction time does not mean that trading volume will increase year over year, but artificial intelligence and automated trading may be important factors driving the expansion of on-chain activity. Investors are still mainly evaluating the value of relevant platforms based on the current size of the crypto market, ignoring the potential market expansion brought about by tokenization. For example, the decentralized trading platform Uniswap may expand its scope of services from crypto assets to traditional asset markets such as stocks, bonds, and real estate in the future. (The Block) This article is sponsored by GENG, Build Your Fortune on GENG (https://geng.one)

3d agoburnking

Circle CEO Unveils Short-Term Roadmap for Smart Body Functionality

Comparing news, Circle CEO Jeremy Allaire said on X that the company has recently released a short-term roadmap for agentic (agentic) functions and related standards. The roadmap covers a number of capabilities being built, with a focus on expanding support for agents as working sellers, and involving mechanisms such as agent identity, reputation, authentication, and trusted discovery. Allaire pointed out that the plan aims to promote the implementation of standards and functions related to smart devices, so that smart bodies can participate in work transactions and service provision. As a USDC stablecoin issuer, Circle continues to focus on integrating infrastructure such as payments and on-chain identity with emerging smart scenarios.

4d ago

NeoSoul officially launched NeoTrade and launched a general smart trading workbench

According to news, NeoSoul, the largest AI economic infrastructure project in the BSC and 0G ecosystem, announced the official launch of NeoTrade today. NeoTrade is a general smart trading workbench for traders, which allows users to directly define how AI trading agents work and put them into operation. Most of the existing automated trading products offer preset strategies. Traders who want more freedom usually need to develop and configure their own trading agents. NeoTrade puts this capability into its trader-oriented product interface. Once launched, the Agent can process market information and execute trades 24/7. The use of funds follows the rules set in advance by the user, and operations requiring approval are subject to confirmation by the user, and the user can also stop the agent at any time. NeoTrade will continuously display the agent's operating status and actual performance. Traders can adjust the allocation according to market results, so that each result becomes the basis for the next adjustment. NeoSoul co-founder Kaelan said that Agentic Trading's next step is to turn what used to require engineering capabilities to be done into products that traders can directly use. NeoTrade hopes to lower this threshold so that more users with trading decisions can use trading agents. As AI Agents move further from assistive analysis to continuous execution, trading products are also beginning to move from pre-defined strategy tools to configurable agent systems. Who can solve autonomous operation and capital control at the same time is becoming the key to competition for Agentic Trading products.

5d ago