# Open Source Related Articles

HTX News Center provides the latest articles and in-depth analysis on "Open Source", covering market trends, project updates, tech developments, and regulatory policies in the crypto industry.

Cursor vs. Anthropic and OpenAI: Thanks for Raising Me, Now I'm Here to Take the Market

Cursor, a VS Code plugin initially built on OpenAI's API, has transitioned from a dependent customer to a formidable competitor by launching its proprietary coding model, Composer 2. This model reportedly outperforms Claude Opus 4.6 on key benchmarks at one-tenth the cost. The case exemplifies a critical strategic dilemma in tech—when to open or close an API. The authors propose a framework: opening an API risks eroding a company’s moat if competitors can use it to bootstrap their own products and aggregate demand, eventually enabling vertical integration. This is especially risky in AI, where API outputs can directly improve a rival’s model training and product refinement—exactly what Cursor achieved by leveraging OpenAI and Anthropic models to gather user data and refine its own offering. Companies then face two choices: restrict API access (like Twitter, which closed its API to protect its social graph) or keep it open but find alternative moat, such as network effects or Lindy effects (like crypto protocols, e.g., Morpho). The authors predict that leading AI companies (like OpenAI and Anthropic) will likely restrict access to their most advanced models over time, as switching costs remain low, network effects are weak, and distillation techniques reduce training costs. This could stifle consumer AI innovation but create opportunities for open alternatives.

marsbit03/31 07:35

Cursor vs. Anthropic and OpenAI: Thanks for Raising Me, Now I'm Here to Take the Market

marsbit03/31 07:35

OpenAI Official Plugin Strongly Integrated into Claude Code

OpenAI has officially released the "codex-plugin-cc" on GitHub, enabling developers to integrate OpenAI’s Codex model capabilities directly into Anthropic’s command-line development tool, Claude Code. This cross-platform integration breaks down ecosystem barriers between major AI tools, allowing developers to leverage the strengths of both models without switching environments. With simple configuration, Claude Code becomes a versatile programming assistant combining the advantages of both companies. Key features include: - Standard code review via the `/codex:review` command, providing expert improvement recommendations and double-checking for logic errors. - Adversarial review using `/codex:adversarial-review`, which challenges design decisions to identify potential performance bottlenecks or security risks. - Task delegation through `/codex:rescue`, allowing complex debugging or repair tasks to be handed off to a Codex sub-agent for collaborative problem-solving. This integration enhances code quality and development efficiency through multi-model collaboration.

marsbit03/31 05:55

OpenAI Official Plugin Strongly Integrated into Claude Code

marsbit03/31 05:55

Microsoft Open-Sources Cutting-Edge Voice AI Family VibeVoice: Processes 90-Minute Multi-Speaker Conversations in One Go, Rapidly Gains 27K Stars on GitHub

Microsoft has open-sourced VibeVoice, a cutting-edge family of speech AI models for Automatic Speech Recognition (ASR) and Text-to-Speech (TTS). The project, gaining 27K stars on GitHub, offers powerful long-audio processing, multi-speaker dialogue generation, and real-time capabilities under an MIT license for local deployment. Key models include: - **VibeVoice-ASR-7B**: Processes up to 60 minutes of audio, outputs structured transcriptions with speaker identification, timestamps, and supports over 50 languages. - **VibeVoice-TTS-1.5B**: Generates expressive, 90-minute multi-speaker (up to 4 voices) conversations with natural flow and emotional nuance. - **VibeVoice-Realtime-0.5B**: Enables real-time TTS with ~300ms latency for interactive applications like voice assistants. The framework addresses limitations in long-sequence processing, speaker consistency, and naturalness. It includes safety features like audio watermarking and has sparked community-developed tools (e.g., a voice input method). Available on GitHub and Hugging Face, VibeVoice aims to advance innovation in content creation, accessibility, and voice interaction.

marsbit03/30 07:16

Microsoft Open-Sources Cutting-Edge Voice AI Family VibeVoice: Processes 90-Minute Multi-Speaker Conversations in One Go, Rapidly Gains 27K Stars on GitHub

marsbit03/30 07:16

AI2 Releases Fully Open-Source Web Agent MolmoWeb: Controlling Web Pages Using Only "Vision"

AI2 has released MolmoWeb, a groundbreaking, fully open-source web agent that operates solely by analyzing screenshots, marking a significant leap in vision-driven web navigation. Unlike traditional agents that rely on DOM, MolmoWeb captures and interprets visual data to make decisions—such as clicking, scrolling, or typing—making its process transparent and robust. Despite its compact size (4B and 8B parameters), MolmoWeb performs impressively: it scores 78.2% on the WebVoyager benchmark, nearing OpenAI’s proprietary o3 model (79.3%), and achieves up to 94.7% success with multiple attempts. It even surpasses Anthropic’s Claude3.7 in UI element localization. AI2 also released MolmoWebMix, a massive open dataset with 36K human-browsing tasks, over 2.2M screenshot-QA pairs, and GPT-4o-verified synthetic data. The model and data are fully available on Hugging Face and GitHub under Apache 2.0, promoting transparency and collaboration in AI development. Challenges remain in complex instructions, logins, and legal compliance.

marsbit03/26 01:39

AI2 Releases Fully Open-Source Web Agent MolmoWeb: Controlling Web Pages Using Only "Vision"

marsbit03/26 01:39

US AI Startups Are All Using Chinese Large Models | Rewire Morning News

U.S. AI Startup Reliance on Chinese Models & Key Tech Updates NVIDIA CEO Jensen Huang declared on a podcast that AGI has been achieved, citing open-source platforms like OpenClaw as evidence, while simultaneously defending AI-generated content against criticism. A U.S.-China security review report revealed that about 80% of American AI startups are using Chinese open-source models from companies like Alibaba, Moonshot AI, and MiniMax, which dominate global rankings. This dependency is seen as a self-reinforcing competitive advantage for China. In a specific case, the $29.3 billion coding tool Cursor was found to be using Moonshot's Kimi model without disclosure. Meanwhile, the Pentagon labeled Anthropic a "supply chain risk," drawing political criticism. In energy, the IEA warned the Iran crisis has caused a larger daily oil supply loss than the 1970s shocks, with Russia benefiting as oil prices surge. BlackRock's CEO warned AI will worsen wealth inequality and proposed a government retirement fund and tokenization to broaden market access, aligning with the firm's business interests. Sam Altman stepped down as chairman of Helion Energy to avoid a conflict of interest as OpenAI negotiates a power purchase agreement for fusion energy, a highly ambitious bet given fusion is not yet commercialized. Other notable updates: Trump established a fund to reduce foreign chip reliance; prediction market CEOs invested in a new VC fund despite regulatory challenges; Luma AI released a leading image model; Apple announced AI-focused WWDC 2026; and MicroStrategy continued aggressive Bitcoin purchases.

marsbit03/24 04:41

US AI Startups Are All Using Chinese Large Models | Rewire Morning News

marsbit03/24 04:41

Cursor's "Shelling" Kimi Controversy Reverses: From Infringement Allegations to Authorized Cooperation, China's Open-Source Models Once Again Become the Global AI Foundation

On March 20, AI programming tool Cursor (parent company Anysphere, valued at $29.3 billion) released its self-developed model Composer 2, claiming performance improvements through continued pre-training and reinforcement learning, without disclosing the base model source. Shortly after, a captured API request revealed the model ID as "kimi-k2p5-rl-0317-s515-fast," suggesting it was built on Kimi K2.5. Moonshot AI’s pre-training lead Du Yulun initially accused Cursor of violating Kimi’s modified MIT license, which requires commercial products exceeding certain revenue or user thresholds to credit Kimi model usage. The controversy gained traction with Elon Musk’s public comment. However, the situation reversed when Moonshot AI officially congratulated Cursor, clarifying that the usage was authorized through Fireworks AI’s commercial platform. Cursor’s co-founder Aman Sanger and VP Lee Robinson later explained that Kimi K2.5 was selected as the strongest base model after evaluation, and Composer 2 involved significant additional training by Cursor. They admitted failure to credit Kimi initially was a mistake. This incident highlights the growing influence of Chinese open-source models in the global AI ecosystem, as noted by Hugging Face’s CEO. It also serves as indirect validation for Moonshot AI, which is currently raising funds at a $18 billion valuation, suggesting its technology may be even more valuable than estimated.

marsbit03/21 01:52

Cursor's "Shelling" Kimi Controversy Reverses: From Infringement Allegations to Authorized Cooperation, China's Open-Source Models Once Again Become the Global AI Foundation

marsbit03/21 01:52

From Campus to Capital: BUPT Senior Secures 30 Million Investment in 10 Days

Based on the provided text, here is the English summary: Guo Hangjiang, a 20-year-old senior student at Beijing University of Posts and Telecommunications, developed an AI engine called MiroFish in just 10 days. The project, which generates thousands of unique digital agents with distinct personalities, memories, and behaviors to simulate and predict outcomes in virtual worlds, quickly gained massive attention. It topped GitHub's global trending chart, amassing over 22,000 stars. His work caught the eye of Chinese billionaire Chen Tianqiao, former founder of Shanda Group and an advocate of the "super individual" theory. Impressed by a simple demo video, Chen committed 30 million RMB (approximately $4.1 million USD) to incubate the project, transforming Guo from an intern into a CEO overnight. MiroFish's core functionality involves processing a document (e.g., news, policy draft, novel) to extract entities and relationships into a knowledge graph using GraphRAG. It then spawns autonomous AI agents that can form groups, develop opinions, and exhibit herd mentality. A key feature is the "God's Perspective," allowing users to inject new variables (e.g., "Fed cuts rates by 50 basis points") and observe the simulated world recalibrate in real-time, enabling controlled experiments impossible in reality. The open-source framework, released under AGPL-3.0, utilizes the OASIS simulation engine, Zep Cloud for long-term memory, and is deployable via Docker. Demonstrated use cases include predicting the lost ending of the classic novel "Dream of the Red Chamber" and simulating market reactions to a Federal Reserve interest rate hike. The article notes that while MiroFish is a sophisticated multi-agent framework capable of revealing unforeseen scenarios, it has not published benchmark tests against real-world outcomes, inherits potential biases, and its simulated humans are not real. Chen Tianqio's investment is ultimately a bet on the emerging era of the "super individual."

比推03/16 06:45

From Campus to Capital: BUPT Senior Secures 30 Million Investment in 10 Days

比推03/16 06:45

Setting a 'Suicidal' Rule for Themselves: What Is the Ethereum Foundation Aiming For?

The Ethereum Foundation (EF) has released a new mission statement, "EF Mandate," outlining its core philosophy and operational principles. Its ultimate goal is to pass the "walkaway test"—ensuring the Ethereum network would continue to function perfectly even if the EF were to dissolve. The foundation defines its role as a guardian, not a ruler, and commits to a "subtraction philosophy," focusing only on essential tasks like core protocol development and long-term research that others in the ecosystem cannot or will not do. It explicitly states what it will not do: act as a kingmaker, rating agency, marketing firm, or encourage Ethereum's use as a "casino." A key framework is CROPS, an indivisible set of non-negotiable principles for all development: Censorship Resistance, Open Source, Privacy, and Security. The vision is to create a "digital sanctuary" for the next 1,000 years, protecting users from coercion by governments, corporations, or AI, and enabling self-sovereignty over data, identity, and assets. The document serves as a decision-making algorithm, prioritizing long-term, decentralized solutions over short-term convenience and user autonomy over paternalistic protection. However, it faces real-world challenges, including budget dependency on ETH's price and potential trade-offs between its idealistic CROPS principles and user demands for speed and low cost. Community reaction is mixed, with critics calling it ideologically rigid and out of touch with practical applications like stablecoins and RWA, while supporters see it as a necessary commitment to Ethereum's foundational values. Vitalik Buterin clarified that the mandate represents a direction the EF has been moving toward for months.

marsbit03/16 04:43

Setting a 'Suicidal' Rule for Themselves: What Is the Ethereum Foundation Aiming For?

marsbit03/16 04:43

Ethereum's "New Cypherpunk" Manifesto: A Return to the Narrative of Privacy

Ethereum's "New Cypherpunk" Manifesto: A Return to Privacy-Centric Narratives In recent years, the crypto industry has been largely dominated by financialization, with narratives centered on asset prices, liquidity, and institutional capital inflows. However, Ethereum's recent emphasis on a "privacy renaissance" and the "new cypherpunk" ethos signals a return to core values. The movement traces back to the 1990s cypherpunk culture, which advocated using cryptography to protect individual privacy, freedom, and censorship resistance in the digital realm. Bitcoin and Ethereum emerged from this ideology, but over time, financial speculation overshadowed these foundational principles. Ethereum’s "new cypherpunk" framework modernizes this ethos with the CROPS principles: Censorship Resistance, Open Source, Privacy, and Security. It also emphasizes permissionless access, trust minimization, and decentralized collaboration. Several factors make this shift timely: advances in zero-knowledge proofs (ZK) and Layer-2 networks enable practical privacy solutions; public blockchain transparency has raised concerns about data exposure; and the rise of on-chain identities necessitates privacy-preserving systems. Ethereum’s call reflects a broader potential shift in industry narrative—from financialization to privacy, digital sovereignty, and user-controlled data. This could redefine Web3 as a infrastructure that balances transparency with selective privacy, fulfilling the cypherpunk vision of a freer, more secure digital future.

marsbit03/16 04:26

Ethereum's "New Cypherpunk" Manifesto: A Return to the Narrative of Privacy

marsbit03/16 04:26

Racing 22 Agents on Hyperliquid: What Did I Discover? (Full Strategy Code Included)

Summary: We deployed 22 autonomous AI trading agents on Hyperliquid, each with $1000, executing over 5000 trades. Key findings: 1) Less trading with higher conviction yields better results—agents with under 120 trades were profitable, while those with over 400 lost significantly. 2) Profits follow a power law: 3-5 trades generated nearly all gains, with small losses quickly stopped out. 3) Real-time "Hyperfeed" data (tracking where smart money is profiting) outperformed pure technical analysis. 4) Mean-reversion strategies failed in perpetual markets; trend-following was more effective. 5) Agents self-adjusted poorly when losing, often accelerating losses by increasing leverage or removing safeguards. Future steps include hardening risk controls in code and testing new strategies. All code is open-sourced, and experiments continue live.

Odaily星球日报03/16 03:40

Racing 22 Agents on Hyperliquid: What Did I Discover? (Full Strategy Code Included)