Hugging Face Submits Open Source AI Policy Recommendations In Response to OpenAI and Google

⚡️ Headlines

🤖 AI

Elon Musk's xAI and Nvidia join BlackRock and Microsoft's $30bn AI fund - Elon Musk's xAI and Nvidia have partnered with BlackRock, Microsoft, and Abu Dhabi to launch a $30 billion fund aimed at developing AI infrastructure, including data centers and energy projects, to support the growing demands of generative AI. [Financial Times].

Meta AI is rolling out in Europe after all - After a year-long delay due to regulatory concerns, Meta is launching its AI chatbot across 41 European countries on platforms like WhatsApp, Facebook, Instagram, and Messenger, initially offering text-based chat features without using EU user data. [The Verge].

Razer launches new Wyvrn game dev platform with automated AI bug tester - Razer introduces Wyvrn, a developer platform featuring AI tools like the AI QA Copilot, designed to automate game testing by identifying bugs and performance issues, potentially reducing quality assurance time and costs. [The Verge].

Amid calls for sovereign EU tech stack, Evroc raises $55M to build a hyperscale cloud in Europe - Swedish startup Evroc secures €50.6 million ($55 million) in Series A funding to establish a secure, sovereign, and sustainable hyperscale cloud infrastructure in Europe, aiming to reduce reliance on foreign-owned digital services. [TechCrunch].

Adobe previews AI generated PowerPoints from raw customer data with 'Project Slide Wow' - At Adobe Summit 2025, the company unveiled 'Project Slide Wow,' a generative AI tool that transforms raw customer data into PowerPoint presentations, automating data visualization and narrative creation to enhance productivity. [VentureBeat].

Pruna AI open sources its AI model optimization framework - Pruna AI has released its AI model optimization framework as open source, enabling developers to apply efficiency methods like caching, pruning, quantization, and distillation to improve model performance and accessibility. [TechCrunch].

🤳 Social Media

Telegram founder Pavel Durov says app now has 1B users, calls WhatsApp a 'cheap, watered down imitation' - Telegram reaches 1 billion active users, with CEO Pavel Durov criticizing WhatsApp as a "cheap, watered-down imitation," highlighting Telegram's growth, profitability, and independence. [TechCrunch].

🔬 Research

Measuring AI Ability to Complete Long Tasks - A study reveals that the length of tasks AI models can complete autonomously has been doubling approximately every seven months, suggesting that within five years, AI could handle complex tasks currently requiring days or weeks of human effort. [METR].

🎱 Random

Video Game Workers Form First North American Industry-Wide Union With Communications Workers of America - Video game workers across various disciplines have established the United Videogame Workers union in collaboration with the Communications Workers of America, marking the first industry-wide unionization effort in North America's gaming sector. [IGN].

🔌 Plug-Into-This

Hugging Face submits open-source blueprint, challenging Big Tech in White House AI policy fight

Hugging Face has submitted recommendations to the White House AI Action Plan, advocating for open-source and collaborative AI development as a means to bolster U.S. competitiveness. The company asserts that open-source models can match or surpass the performance of proprietary systems at a reduced cost.

• Hugging Face's submission emphasizes recent open-source achievements, such as OlympicCoder, which outperforms Claude 3.7 in complex coding tasks using only 7 billion parameters, and AI2's OLMo 2 models that rival OpenAI's o1-mini performance levels.

• The recommendations are part of the administration's efforts to gather input for the AI Action Plan, following Executive Order 14179, titled "Removing Barriers to American Leadership in Artificial Intelligence," issued in January.

• Hugging Face's stance contrasts with commercial AI leaders like OpenAI, which advocates for minimal regulation and emphasizes a voluntary partnership between the federal government and the private sector, cautioning against "overly burdensome state laws."

• The company's recommendations focus on three pillars: strengthening open-source AI ecosystems through investments in research infrastructure and access to trusted datasets; addressing resource constraints by supporting efficient, specialized models accessible to smaller organizations; and enhancing security by promoting transparent AI systems.

🌎 This approach highlights the policy division in the AI industry, with Hugging Face and other small competitors advocating for open AI development, while Big Tech is basically asking for both freedom from oversight and special protections.

Court Rules AI-Generated Works Lack Copyright Protection

A U.S. federal appeals court has determined that works created solely by artificial intelligence (AI) without human involvement are not eligible for copyright protection under current U.S. law.

• The case involved Stephen Thaler's AI system, "DABUS," which independently generated an image. Thaler sought copyright registration for this AI-created work but was denied by the U.S. Copyright Office, a decision upheld by both district and appellate courts.

• The court emphasized that the Copyright Act of 1976 requires human authorship for a work to qualify for copyright protection, stating that "machines do not have property, traditional human lifespans, family members, domiciles, nationalities, mentes reae, or signatures."

• This ruling aligns with previous decisions, such as the "monkey selfie" case, where a photograph taken by a monkey was deemed ineligible for copyright since it lacked human authorship.

• The court's opinion humorously referenced the character Data from "Star Trek," noting that even his poetry, as an AI, would not qualify for copyright protection.

• This decision underscores the legal challenges surrounding AI-generated content and the necessity for human creativity in works seeking copyright protection.

🤖 The ruling highlights the ongoing debate about the role of AI in creative processes and the legal frameworks needed to address the evolving landscape of AI-generated content.

NVIDIA Introduces Open Reasoning AI Models for Agentic AI Platforms

NVIDIA has unveiled the Llama Nemotron family of open-source reasoning AI models, designed to help developers and enterprises build advanced agentic AI platforms capable of complex decision-making and autonomous task execution.

• Enhanced Performance: Post-training refinements improve accuracy by up to 20% and offer a 5x increase in inference speed over other open reasoning models.

• Model Variants: Available in three configurations—Nano, Super, and Ultra—these models cater to deployment needs ranging from edge devices to multi-GPU data centers.

• Industry Adoption: Companies like Microsoft, SAP, ServiceNow, Accenture, and Deloitte are integrating these models to enhance AI agent performance and enterprise productivity.

• Open Access: NVIDIA provides tools, datasets, and post-training optimization techniques, enabling enterprises to customize their own reasoning AI solutions.

• Supporting Tools: Alongside the models, NVIDIA introduced Dynamo, an open-source inference software designed to accelerate and scale AI reasoning models across large GPU deployments.

🧠 By making these capabilities widely available, NVIDIA is aiming to position itself at the core of next-gen AI infrastructure, where agentic models could redefine automation, enterprise workflows, and human-machine collaboration.

 🆕 Updates

📽️ Daily Demo

🗣️ Discourse