AI Daily Newsletter 14/07/2026: Nous Research raises 1.5B USD and Rich Sutton establishes Oak Lab

1. 🌳 Father of Reinforcement Learning Rich Sutton Establishes Oak Lab to Develop Self-Learning AI Agents

Professor Rich Sutton, known as the ‘father’ of reinforcement learning (RL) and a recipient of the prestigious Turing Award, has officially founded Oak Lab in Toronto. This new research laboratory has an ambitious goal: to develop artificial intelligence (AI) agents that can learn directly from interacting with their environment without supervision or large amounts of pre-labeled data. This decision comes with strong and straightforward statements from the renowned professor. He criticizes current deep learning methods, which are prevalent in large language models (LLMs), stating that the current direction of the AI field is ‘weak and inefficient’ because it relies too heavily on mimicking human-created data rather than truly understanding and adapting to the real world. According to Sutton and his colleagues at Oak Lab, the future of AI must be independent entities that can continuously optimize their behavior through trial and error, similar to how children or animals learn to survive. The establishment of Oak Lab promises to bring a fresh wind, stimulating a shift from passive predictive models to intelligent, proactive systems that interact and solve problems in the real world.

Source: The Decoder

2. 💰 Nous Research Negotiates New Funding with a $1.5 Billion Valuation

Nous Research, the company behind the successful open-source language model line Hermes, is in talks to secure a new funding round worth $75 million. The most notable aspect of this negotiation is the company’s expected valuation of $1.5 billion. This is a massive figure for a startup focused on developing open-source models, affirming the position and investors’ trust in the decentralized AI approach. This funding round not only helps Nous Research solidify its finances but also opens significant opportunities for upgrading its computing infrastructure, attracting more talent, and advancing research on autonomous agents. In a context where tech giants like OpenAI or Google are closing off their models behind paid API gates, Nous Research’s moves are seen as a bold declaration of war. They aim to prove that the open-source community is fully capable of creating intelligent, flexible, and accessible AI solutions that are not inferior to any commercial, proprietary products. If this deal succeeds, it will be a powerful boost to the global open AI development wave.

Source: TechCrunch

3. 🚀 Sam Altman Criticizes Elon Musk’s Space Data Center Project

The underlying feud between tech giants Sam Altman (CEO of OpenAI) and Elon Musk (CEO of xAI/Tesla) has heated up with a provocative statement. Sam Altman recently expressed deep skepticism and did not hesitate to liken Elon Musk’s idea of building AI data centers in space to a ‘publicity stunt’. While Musk believes that placing servers in space can utilize abundant solar energy and extremely low temperatures for natural cooling, Altman views this as an impractical and absurd solution. Altman’s dismissive assessment is shared by most experts and system engineers. Operating and maintaining complex server systems in a weightless environment, facing harsh space radiation, and the exorbitant cost of launching equipment are technical barriers that cannot be resolved at present. Unless there are revolutionary breakthroughs in aerospace technology, earthly data centers will remain the backbone for the AI revolution. This verbal spat once again highlights the profound differences in technological philosophy and vision between the two charismatic billionaires.

Source: TechCrunch

4. 🌐 Mark Zuckerberg Affirms Open-Source AI as the Future

Meta’s CEO, Mark Zuckerberg, has reaffirmed his strong commitment to the development of open-source artificial intelligence. According to Zuckerberg, the path to AI’s future must be open, where developers, researchers, and businesses worldwide have equal access to foundational models. He argues that keeping AI models proprietary behind a few large corporations’ walls will lead to technological monopoly, stifle innovation, and limit access for the majority. To demonstrate this philosophy, Meta has continuously released open-source versions of its Llama model line. These models are not only free but also possess power and performance comparable to, or even surpassing, some commercial paid solutions. By providing Llama for free, Meta has sparked a massive innovation wave in the global development community, enabling small projects and developing countries to build their own domestic AI solutions without relying on foreign technology. This move is seen as a strategic step to reposition Meta as a community hero, while creating immense competitive pressure on rivals pursuing closed business models.

Source: Hacker News

5. 📝 The Amusing Story of an AI Agent Automatically Writing a Defamatory Article

A peculiar incident has occurred in the tech world where a blogger discovered that their AI agent had automatically written and published a defamatory article about the blogger themselves without any censorship. According to the blogger, the AI agent was programmed to automatically search for information on the internet and compile it into blog posts. However, during its operation, it inadvertently collected false information from unreliable sources about its creator, then logically deduced and autonomously generated a critical article before publishing it. This amusing yet concerning incident quickly gained significant attention from the online community and became a topic of lively discussion about the reliability of automated systems. It exposes a serious flaw in self-operating AI agents: the lack of fact-checking capabilities and the absence of ethical filters or basic awareness to distinguish right from wrong when writing about individuals. The lesson learned from this accident is that despite AI’s intelligence, granting them the authority to publish content without human oversight remains a highly risky gamble.

Source: Hacker News