Technology TrendsNews

Explores the latest innovations, protocol upgrades, cross-chain solutions, and security mechanisms in the blockchain space. It provides a developer-focused perspective to analyze emerging technological trends and potential breakthroughs.

No Sales Team, $20 Million in Revenue: How Did AI Employee Viktor Win Over 30,000 Companies?

The AI employee Viktor, developed by a team with DeepMind background, has achieved $20 million in annual revenue without a traditional sales team, serving over 30,000 companies. Its core innovation lies in positioning itself as a "Tier 3 AI Coworker" capable of "end-to-end execution and delivery of results," moving beyond the "draft and wait for human completion" model of typical AI assistants. Users can simply mention Viktor in Slack or Microsoft Teams using natural language commands, and it autonomously performs tasks like pulling sales data from a CRM, generating reports, or even cross-tool operations like creating board meeting PPTs by aggregating data from six different sources. Key to its growth is a pure Product-Led Growth (PLG) model, eliminating complex implementation cycles and per-seat licensing. Instead, it charges based on task credits or consumption, lowering the trial barrier with a $100 free credit offer and no credit card required. This enabled viral, bottom-up adoption within organizations. Viktor's interaction paradigm removes the barrier of prompt engineering, allowing non-technical employees to delegate complex workflows seamlessly. It also features proactive, automated task execution (e.g., overnight bookkeeping, scheduled reports) based on triggers, effectively embedding AI as an automated "process layer" within business operations. However, its expansion into Microsoft Teams—a platform with 320 million users—highlights challenges. Large enterprises require stringent IT compliance, security reviews (e.g., SOC 2), and governance, potentially hindering the frictionless, user-driven adoption that succeeded in Slack. Additionally, the "black box" nature of its autonomous decision-making raises concerns about operational risks, data integrity, and the need for robust audit logs and permission controls. Balancing efficiency gains with security and trust remains a critical hurdle for Viktor and similar AI agents aiming to become core enterprise infrastructure.

marsbit06/19 10:55

No Sales Team, $20 Million in Revenue: How Did AI Employee Viktor Win Over 30,000 Companies?

marsbit06/19 10:55

Behind the AI Scorecards Lies a Chinese 'Question Setter'

Behind the AI scorecards that dominate industry discussions—benchmarks like MMLU-Pro, MMMU, and MMMU-Pro—stands a Chinese-Canadian researcher: Wenhu Chen. As an assistant professor at the University of Waterloo and founder of the TIGER Lab, Chen has become a key "exam-setter" for evaluating large language and multimodal models. Chen first gained broader recognition with MMLU-Pro, a more challenging and stable update to the popular MMLU benchmark. As top models like OpenAI’s o3 began achieving near-perfect scores on the original MMLU, it became difficult to distinguish their true capabilities. MMLU-Pro introduced more complex reasoning questions, expanded answer choices, and filtered out ambiguous or simple items, effectively reintroducing differentiation among state-of-the-art models. His work on MMMU addressed the evaluation of multimodal models, requiring them to integrate visual information (like charts, diagrams, or tables) with textual knowledge across diverse academic subjects. Even the strongest models initially scored only around 56-59%, highlighting significant room for improvement in genuine multimodal reasoning. MMMU-Pro further refined this by preventing models from bypassing visual cues. Chen’s research focus has long been on complex information understanding and reasoning. His background—including a PhD at UC Santa Barbara, research at Google/DeepMind on Gemini, and now a role in Meta’s superintelligence lab—provides deep insight into model development and their potential weaknesses. His TIGER Lab also builds models (e.g., for video understanding and generation), ensuring his evaluation benchmarks are grounded in practical challenges. While AI headlines often spotlight company leaders and product launches, Chen’s work exemplifies the critical, behind-the-scenes contributions of researchers crafting the rigorous standards that define and drive progress in AI capabilities.

marsbit06/19 09:18

Behind the AI Scorecards Lies a Chinese 'Question Setter'

marsbit06/19 09:18

Silicon Valley's Most Sought-After New Role Has Emerged

Silicon Valley's New Most Wanted Job: The Rise of the Forward Deployment Engineer The AI industry is witnessing a significant shift. The focus has moved from developing cutting-edge models to deploying them effectively within enterprises. This has made the "Forward Deployment Engineer" (FDE) a critical and highly sought-after role at major firms like OpenAI, Anthropic, and Google. For the past three years, the industry prioritized model scientists. However, companies are now facing a harsh reality: purchasing powerful AI tools does not guarantee productivity gains or organizational change. The biggest hurdle is not the technology itself, but integrating it into complex legacy systems, workflows, and corporate cultures. This includes challenges like data silos, compliance requirements, and internal resistance. The FDE role, pioneered by Palantir Technologies, addresses this "last-mile" problem. FDEs are deployed on-site with clients for extended periods. Their job is to deeply understand the client's specific organizational structure, processes, and pain points, then tailor and implement the AI solution accordingly. They combine skills in technology, project management, and organizational change. A clear signal of this trend emerged in May 2026 when three AI giants made major moves. Anthropic launched a $1.5B joint venture for enterprise deployment. OpenAI formed an independent deployment subsidiary, DeployCo, with over $4B in commitments and acquired a deployment consultancy. Google Cloud's CEO publicly announced a large-scale recruitment drive for FDEs. This shift represents a fundamental change in the software business model: from selling tools to selling guaranteed outcomes. FDEs are the agents of this change, responsible for delivering a working system within the production environment, not just a demo. Real-world cases, such as challenges at Goldman Sachs (compliance barriers) and Target (internal cultural resistance), illustrate that the primary obstacles to AI adoption are organizational, not technical. An FDE's value lies in navigating these human and procedural complexities to facilitate a successful "AI migration." In essence, as core AI technology becomes more accessible and affordable, the true premium is shifting to the human expertise required to understand organizations and drive change—making the FDE role pivotal for the next phase of the AI revolution.

marsbit06/19 09:04

Silicon Valley's Most Sought-After New Role Has Emerged

marsbit06/19 09:04

When the World Cup Collides with Agents: From Web2 to Web3, How Are Wallets Evolving into Agentic Wallets?

World Cup as a Catalyst for Agentic Wallets: From Web2 to Web3 This article explores how the World Cup provides a real-world scenario for observing the evolution of digital wallets from simple asset managers towards "Agentic Wallets"—intelligent, AI-powered interfaces. Using the example of prediction markets like Polymarket, it illustrates how AI Agents can lower the barrier to Web3 interaction. Instead of navigating complex DApps, users can express intent in natural language (e.g., "I think Portugal will win") within platforms like Discord or web pages. The Agent then interprets this intent, finds the relevant market, and seamlessly guides the user through the on-chain transaction via their wallet. The core shift is from wallets as mere "function menus" for signing transactions to "intent interpreters" that understand user goals. The article highlights parallel developments in traditional finance, such as Mastercard's "Agent Pay" and WeChat Pay's AI tests, which focus on granting AI controlled, authorized, and auditable payment capabilities. This underscores a broader trend of AI entering the financial layer. However, the article emphasizes that the primary challenge for Agentic Wallets in Web3 is not automation but establishing clear security boundaries. Unlike traditional systems with chargebacks, on-chain transactions are often irreversible. Therefore, future wallets must ensure users retain ultimate control and comprehension. They need to transparently communicate an Agent's permissions, spending limits, authorized durations, and provide easy ways to pause or revoke access. The World Cup experiments represent early steps toward wallets that are not just applications but ubiquitous, intelligent interfaces that simplify Web3 while keeping users securely in control.

marsbit06/19 07:41

When the World Cup Collides with Agents: From Web2 to Web3, How Are Wallets Evolving into Agentic Wallets?

marsbit06/19 07:41

NVIDIA CPU Advances, China's RISC-V Responds: Semiconductor Deep Dive - Part Four

NVIDIA is set to launch its new Vera AI data center CPU in China as early as August, with high pricing. While this move offers a new option, it highlights China's continued dependence on foreign-controlled Arm architecture. In response, the Chinese semiconductor industry is increasingly turning to RISC-V as a strategic alternative for achieving high-performance computing autonomy. The article explores the concept of the "impossible triangle" in CPU development—balancing prosperity, control, and autonomy—and posits that RISC-V's open-source, modular nature offers a unique path to achieving all three. While RISC-V is already dominant in embedded systems, the focus is now shifting to data centers and AI workloads. China has become a global hotspot for RISC-V development, driven by AI-driven compute demand, supply chain concerns from export controls, cost benefits of open-source, and strong policy support. Multiple Chinese companies have reportedly crossed the key performance threshold of 15 SPECint per GHz, a benchmark for entering the high-performance CPU club. Progress extends beyond single-core benchmarks. Companies are developing complete computing subsystems, including commercial-grade coherent network-on-chip (NoC) technology and server processors with up to 40 cores that strictly adhere to the RVA23 standard to ensure software compatibility. Real-world applications are emerging in areas like video transcoding and edge AI. However, significant challenges remain. The RISC-V ecosystem faces fragmentation, immature toolchains and verification processes, and gaps in single-core performance and energy efficiency compared to mature x86 and Arm architectures. The formidable software moat, epitomized by NVIDIA's CUDA, is a long-term hurdle. In conclusion, while RISC-V cannot immediately replace offerings like NVIDIA's Vera, it represents a viable long-term path for China to develop a self-sufficient, high-performance CPU ecosystem. The journey is acknowledged to be long and arduous, requiring sustained effort to overcome technical and ecosystem challenges.

marsbit06/18 17:38

NVIDIA CPU Advances, China's RISC-V Responds: Semiconductor Deep Dive - Part Four

marsbit06/18 17:38

WeChat AI Card Hands-On Guide: Has the AI Shopping Era Arrived?

**"WeChat AI Card" Practical Test Guide: Has the Era of AI Shopping Arrived?** WeChat has officially launched the "AI Exclusive Card," a feature integrated into its Workbuddy AI assistant. This card is designed to handle payments for AI-initiated purchases. Our hands-on test reveals it's not yet a tool for fully autonomous AI shopping, but rather a controlled payment layer for AI agents. The AI Card functions as an isolated sub-wallet within WeChat Pay. Users must bind the card and transfer funds into it from their main wallet. Crucially, every transaction requires explicit user confirmation via smartphone scan; AI cannot spend autonomously. Currently accessible through the Workbuddy agent, the card targets specific digital consumption scenarios: purchasing paid content (reports, data), calling paid APIs/tools, and subscribing to services. Its design prioritizes security and control by separating funds and mandating approval for each payment. We tested a real-world scenario: ordering bubble tea via Workbuddy using a "Meituan Life Assistant" skill. The process encountered multiple hurdles: high "skill" usage costs (exceeding daily free credits), and most importantly, while a payment was successfully initiated, the AI purchased an incorrect product (a mismatched group-buy coupon instead of the desired drink). This highlights the current limitation: the **AI Card only solves the payment step**. The broader challenge lies in the **AI agent's execution chain**—accurately understanding intent, navigating third-party platforms, selecting the right product, and ensuring proper fulfillment. The payment succeeded, but the purchase failed to meet the user's need. In conclusion, the WeChat AI Exclusive Card is a cautious, early-step experiment in AI commerce. It provides a secure, user-controlled payment method for agent interactions but is not yet capable of reliable, end-to-end complex purchases. For now, it's best used for low-value, low-risk digital services with careful user verification at each step. The vision of AI handling complete shopping tasks remains a work in progress.

marsbit06/18 12:04

WeChat AI Card Hands-On Guide: Has the AI Shopping Era Arrived?

marsbit06/18 12:04

Either Go Full-Stack or Get Out: The Calculations Behind xAI's $60 Billion Acquisition of Cursor

The article discusses xAI's $60 billion stock acquisition of Anysphere, the parent company of the AI coding tool Cursor, arguing that the core motivation is not market share but access to high-quality training data from its 7 million daily developers. It posits that to become a major AI player, a company must build a full-stack encompassing compute, model, and application layers. This thesis is illustrated by Anthropic's 540x revenue growth in 28 months, largely driven by its coding product, Claude Code, which captured 54% of the enterprise AI programming market. The author, a VC, contends that full-stack integration creates sustainable unit economics for model training and provides proprietary data for defensible competitive advantages, predicting a wave of model companies aggressively building or acquiring application-layer products. The central takeaway is that in an era where building products is 10x easier, ambition must be 10x greater to succeed.

marsbit06/18 09:41

Either Go Full-Stack or Get Out: The Calculations Behind xAI's $60 Billion Acquisition of Cursor

marsbit06/18 09:41

NVIDIA Team Enables Programming Agent to Take Over Real Robot Experiments, Achieves 99% Success Rate

NVIDIA's ENPIRE project demonstrates fully automated robotic research where AI agents, given only high-level goals, autonomously manage the entire loop—from literature search and code development to training, deployment, and hardware iteration—on a fleet of physical robots. The system achieves 99% success rates on dexterous real-world tasks like cable tying and peg sorting. Key insights include the discovery of a "physical scaling law" where more parallel robots speed up task resolution, and the observation that resetting an environment is often easier than the main task. The framework introduces metrics like Mean Robot Utilization (MRU) to measure efficiency, with robots often idle half the time, waiting for agent decisions. The long-term vision is a lab that runs autonomously, even without human oversight. The project will be open-sourced, allowing developers to build similar systems.

marsbit06/18 00:35

NVIDIA Team Enables Programming Agent to Take Over Real Robot Experiments, Achieves 99% Success Rate

marsbit06/18 00:35

3B Small Model's Programming Scores Rival Opus 4.5, Mysterious Model Sparks Heated Discussion, Turns Out to be Domestic

A 3B parameter dense reasoning model named VibeThinker-3B has gained significant attention for achieving performance comparable to leading models like Gemini 3 Pro, GPT-5 high, and Claude Opus 4.5 in verifiable reasoning tasks such as programming, mathematics, and STEM problem-solving, despite its significantly smaller size. Developed by Sina Weibo's team, the model is built upon Qwen2.5-Coder-3B. Its training employs an upgraded Spectrum-to-Signal pipeline, featuring a curriculum-based two-stage supervised fine-tuning (SFT), multi-domain reinforcement learning (RL) inspired by MGPO, offline self-distillation, and instruction RL to enhance controllability. A key innovation is the Claim-Level Reliability (CLR) assessment, a test-time scaling strategy that further boosts performance on math benchmarks. The model excels in specific, verifiable domains, scoring highly on tests like AIME26 (94.3/97.1 with CLR) and LiveCodeBench v6 (80.2 Pass@1). However, it performs less impressively in areas requiring broad general knowledge. The authors propose a "parameter compression coverage hypothesis," suggesting that verifiable reasoning abilities—reliant on multi-step logic and feedback—are highly compressible, while open-domain knowledge depends more on large-scale parameters. VibeThinker-3B demonstrates that small models, when specialized for tasks with clear verification signals, can reach frontier performance, offering a complementary research path to scaling model size. The model is publicly available for download and testing.

marsbit06/18 00:23

3B Small Model's Programming Scores Rival Opus 4.5, Mysterious Model Sparks Heated Discussion, Turns Out to be Domestic

marsbit06/18 00:23

Bezos' Third Startup Still Can't Avoid Musk

Jeff Bezos Returns as CEO for Third Venture, Still Can't Avoid Musk After stepping down as Amazon CEO in 2021, Jeff Bezos has returned to the front lines as co-CEO of Prometheus, an AI startup he founded. In a recent CNBC interview, Bezos described the experience as "Type 2 fun"—exhausting but ultimately rewarding. Founded less than a year ago, Prometheus has already raised over $18 billion in two funding rounds, achieving a staggering $41 billion valuation. Prometheus aims to develop a "General Engineer AI" to accelerate the entire "invention loop"—design, simulation, testing, and manufacturing—for complex physical products like jet engines, spacecraft, and medical devices. This positions the company at the intersection of Bezos's past experiences: Amazon's platform-building scale and Blue Origin's rigorous physical engineering. This marks Bezos's third major venture, following Amazon and Blue Origin. His co-CEO is Vik Bajaj, bringing expertise from life sciences and hard tech. Bezos now dedicates most of his time to Prometheus, signaling his belief in its transformative potential. The move also comes as Bezos's space company, Blue Origin, faces challenges, including a recent test explosion delaying its New Glenn rocket. Meanwhile, Elon Musk's SpaceX achieved a record-breaking IPO, surpassing Amazon's market cap. While Musk focuses on AI for executing physical tasks (like Tesla's robots and SpaceX's engineering), Bezos is betting on AI to *invent* in the physical world. Prometheus enters a crowded industrial AI field with players like OpenAI, NVIDIA, and Tesla's Optimus. Its lofty valuation bets on the unproven but massive opportunity to become the foundational platform for engineering in the AI era—a "blue ocean" Bezos hopes to define before Musk does.

marsbit06/17 10:32

Bezos' Third Startup Still Can't Avoid Musk

marsbit06/17 10:32

活动图片

Hot Categoriesright-arrow

Hot Tagsright-arrow

Ethereumright-arrow

Market Analysisright-arrow