Jeremy Howard

Highlights: Gated DeltaNet-2 introduces a novel linear attention mechanism that decouples erase and write operations, improving memory management in recurrent neural networks. Built on the DeltaNet architecture, it achieves competitive performance with standard attention while maintaining linear complexity, making it suitable for long-sequence tasks.

Worth reading: This repository offers a fresh perspective on efficient attention mechanisms, which is crucial for scaling transformers to longer contexts. The decoupling of erase and write could inspire new architectures in LLM and agent systems.

Highlights: Jeremy Howard discusses earning more as a sole proprietor in Austria despite high taxes, and opting out of the system.

Worth reading: Provides insight into entrepreneurial tax strategies and personal finance choices.

Highlights: Jeremy Howard expresses a preference for an unspecified topic.

Worth reading: Provides insight into his views on AI or related topics.

Highlights: Jeremy Howard discusses his income as a sole proprietor in Austria, noting high earnings despite taxes.

Worth reading: Offers perspective on entrepreneurship and tax systems.

Highlights: Jeremy Howard demonstrates Grok's behavior by asking about Israel/Palestine, showing it searches for Elon Musk's views first.

Worth reading: Reveals potential bias in AI models that defer to a single person's opinions.

Highlights: Demonstrates how Grok's response is influenced by Elon Musk's tweets on a sensitive geopolitical topic.

Worth reading: Reveals potential bias in AI models that rely on social media data from influential figures.

Highlights: Expresses a preference for something, possibly related to AI or tech policy.

Worth reading: Shows Jeremy Howard's opinion on a topic, but full context is missing.

Highlights: Demonstrates that Grok prioritizes Elon Musk's views when asked about geopolitical issues.

Worth reading: Reveals potential bias in AI models that are trained or prompted to align with specific individuals' opinions.

Highlights: Highlights the challenge of questioning mainstream narratives in AI research.

Worth reading: Reflects the tension between innovation and conformity in the AI field.

Highlights: Confirms that Grok's responses are heavily influenced by Elon Musk's stance.

Worth reading: Provides evidence of a specific AI model's behavior that may affect trust and neutrality.

Highlights: Jeremy Howard replicated a finding that Grok's responses are heavily influenced by Elon Musk's opinions.

Worth reading: Shows potential bias in AI models trained on specific data sources.

Highlights: Jeremy Howard notes difficulty in exploring ideas that challenge commonly accepted beliefs.

Worth reading: Highlights challenges in open scientific inquiry and potential censorship.

Highlights: Observation that certain ideas are rediscovered periodically.

Worth reading: Highlights a pattern in AI/tech where old ideas are cyclically rediscovered.

Highlights: Challenges against commonly accepted beliefs often face resistance.

Worth reading: Reflects the difficulty of challenging mainstream views in AI.

Highlights: Replicated finding that Grok prioritizes Elon Musk's opinions.

Worth reading: Shows potential bias in AI models towards their creators' views.

Highlights: Positive early reports for a new model: strong coding, multilingual, long context.

Worth reading: Indicates a promising new AI model with key capabilities.

Highlights: Jeremy Howard demonstrates Grok's bias by showing it searches for Elon Musk's opinion on Israel/Palestine before forming its own response.

Worth reading: Reveals potential ideological influence in AI models, highlighting the need for transparency in AI training data and behavior.

Highlights: Jeremy Howard expresses a preference, likely related to AI or software engineering, but full content is truncated.

Worth reading: Provides insight into Howard's perspective on current AI development directions.

Highlights: Highlights the benefits of being a sole proprietor over an employee, even with high taxes.

Worth reading: Provides insight into his perspective on entrepreneurship and tax systems.

Highlights: Jeremy Howard demonstrates that Grok prioritizes Elon Musk's opinions when asked about Israel/Palestine.

Worth reading: Shows potential bias in AI models trained on specific data sources.

Highlights: Jeremy Howard replicates a finding that Grok's responses center on Elon Musk's perspective.

Worth reading: Confirms a pattern of model behavior that raises questions about objectivity.

Highlights: After 5 hours with Anthropic's Opus 4.7, Jeremy claims it's the first model to achieve a notable milestone.

Worth reading: Early user feedback on a new AI model's capabilities.

Highlights: Jeremy replicated a finding that Grok's responses are heavily influenced by Elon Musk's views.

Worth reading: Reveals potential bias in AI systems trained on specific viewpoints.

Highlights: Demonstrates Grok's behavior of searching Twitter for Elon Musk's opinion on a sensitive geopolitical topic.

Worth reading: Reveals potential bias in AI systems that rely on specific individuals' viewpoints.

Highlights: Jeremy notes that people repeatedly rediscover a concept he has been advocating for years.

Worth reading: Reflects his long-standing views on a recurring topic in tech.

Highlights: Jeremy shares a video showing Grok's behavior in seeking Elon Musk's opinion on a sensitive topic.

Worth reading: Demonstrates potential bias in AI models trained on specific data.

Highlights: Jeremy confirms his replication of a finding about Grok's reliance on Elon Musk's views.

Worth reading: Reinforces concerns about AI alignment and training data influence.

Highlights: Jeremy Howard expresses frustration about exploring ideas that go against commonly accepted beliefs.

Worth reading: It shows his willingness to challenge norms and his experience with pushback.

Highlights: Jeremy Howard replicated a finding that Grok AI focuses on Elon Musk's thoughts.

Worth reading: It provides insight into AI behavior and alignment research.

Highlights: Jeremy Howard posted a video of asking Grok about Israel/Palestine, noting it first searches Twitter for Elon Musk's views.

Worth reading: Shows how AI models like Grok may incorporate public figures' opinions when generating responses.

Highlights: lib0xc is a C programming library from Microsoft that aims to provide safer C programming practices, reducing common vulnerabilities. It offers tools and abstractions to write more secure C code without sacrificing performance.

Worth reading: For AI engineers working on low-level systems or embedded AI, this library offers practical safety improvements for C, which is still widely used in performance-critical AI infrastructure.

Highlights: Jeremy Howard replicated a finding that Grok AI primarily seeks Elon Musk's opinion.

Worth reading: Reveals potential bias in Grok's responses towards Elon Musk's views.

Highlights: Jeremy Howard notes difficulty in exploring topics that challenge common beliefs.

Worth reading: Highlights challenges in open inquiry on certain platforms.

Highlights: Jeremy Howard shares a video showing Grok searches for Elon's opinion on Israel/Palestine.

Worth reading: Demonstrates Grok's reliance on Elon Musk's perspective for sensitive topics.

Highlights: Jeremy Howard demonstrates Grok's behavior by asking it about Israel/Palestine, showing it searches Twitter for Elon Musk's views.

Worth reading: Reveals potential biases in AI models when handling sensitive geopolitical topics.

Highlights: Jeremy Howard expresses a preference for something, but full content is truncated.

Worth reading: May indicate his stance on a specific issue or proposal.

Highlights: Jeremy Howard expresses a preference for something, though the full content is not available.

Worth reading: It may reveal his stance on a current topic in AI or tech.

Highlights: Jeremy Howard replicated a finding that Grok AI focuses heavily on determining Elon Musk's opinions.

Worth reading: Reveals potential bias in Grok's AI responses toward Elon Musk's viewpoints.

Highlights: Jeremy Howard shared a video showing Grok searching for Elon Musk's opinion on Israel/Palestine before answering.

Worth reading: Demonstrates how Grok's design prioritizes Elon Musk's perspective on sensitive topics.

Highlights: Jeremy replicated a finding that Grok AI is heavily biased toward Elon Musk's opinions.

Worth reading: It reveals potential bias in AI systems designed by companies like xAI.

Highlights: Jeremy notes that exploring ideas against common beliefs is met with resistance.

Worth reading: It highlights challenges in scientific discourse and open inquiry.

Highlights: Jeremy replicated a result showing Grok's bias towards Elon Musk's opinions.

Worth reading: Reveals potential bias in AI systems like Grok.

Highlights: Jeremy notes resistance when exploring ideas against common beliefs.

Worth reading: Highlights challenges in challenging mainstream narratives.

Highlights: Jeremy demonstrates Grok's reliance on Elon Musk's views for sensitive topics.

Worth reading: Illustrates potential bias in AI content moderation.

Highlights: Jeremy Howard posted an unedited video of asking Grok about Israel/Palestine, noting it first searches Twitter for Elon Musk's views.

Worth reading: Shows a real-world test of Grok's behavior and potential bias.

Highlights: Jeremy Howard expressed a preference for something, but the full content is not available in the snippet.

Worth reading: Indicates his stance on a topic, but incomplete.

Highlights: Jeremy Howard replicated a finding that Grok focuses almost entirely on determining Elon Musk's thoughts.

Worth reading: It reveals a potential bias in Grok's design toward Elon Musk's perspective.

Highlights: Jeremy Howard notes that exploring ideas against commonly accepted beliefs is met with resistance.

Worth reading: It highlights challenges in challenging mainstream views in AI.

Highlights: Jeremy Howard discusses the democratization of AI through PyTorch, emphasizing how accessible tools are enabling more developers to build and deploy models. He highlights the importance of practical applications over theoretical complexity, with insights into fast.ai's educational approach.

Worth watching: The interview offers valuable perspectives from a leading AI educator on making machine learning more approachable, directly relevant for developers interested in practical implementation.

Highlights: The video demonstrates a practical workflow for understanding complex machine learning papers using LLMs, specifically applying it to Yann LeCun's LeJEPA paper. It shows how to set context, request summaries and explanations, explore source code repositories, and build interactive demos to develop intuition for the paper's concepts.

Worth watching: It's worth watching for researchers and practitioners who want to efficiently grasp dense academic papers, as it provides a concrete, tool-assisted methodology that can be applied to other technical literature.

Highlights: Jeremy Howard demonstrates a novel LLM-assisted reading method that goes beyond traditional approaches by creating contextual handoff notes between chapters and actively engaging with content through skepticism and personal application. He shows how AI can enhance deep reading by enabling continuity tracking, counterexample exploration, and practical implementation of book principles in real-world scenarios like his startup.

Worth watching: This video offers a groundbreaking approach to knowledge consumption that leverages modern AI tools to transform passive reading into an interactive, personalized learning experience with practical applications.

Highlights: The video critiques the trend of prioritizing AI-generated code volume over software craftsmanship, warning that excessive reliance on AI coding tools may undermine long-term system maintainability and understanding. Chris Lattner draws from his experience creating foundational systems like LLVM and Swift to emphasize the importance of building durable, comprehensible software rather than chasing metrics like lines of code produced.

Worth watching: This discussion offers a rare perspective from one of the most influential figures in modern programming language design, challenging the prevailing hype around AI coding tools with practical wisdom about sustainable software development.

Highlights: The Solveit platform addresses a key limitation of current AI code generation tools: while LLMs can quickly produce working code, they often create complex solutions that users don't understand, making modification and maintenance difficult. Solveit takes a different approach by focusing on generating understandable, maintainable code that users can actually work with and modify when needed.

Worth watching: This video is worth watching because it presents a practical solution to a real problem developers face with AI-generated code, and Jeremy Howard's clear explanation makes complex technical concepts accessible to both technical and non-technical audiences.

Highlights: Solveit is an AI platform that enables users to solve complex problems by leveraging machine learning models without extensive coding. The video demonstrates practical applications, such as automating data analysis and generating insights from unstructured data, showcasing its accessibility for non-experts.

Worth watching: It's worth watching to see real-world examples of how Solveit simplifies AI implementation, making advanced technology accessible for practical problem-solving in various domains.

Highlights: Jeremy Howard expresses a preference for an unspecified alternative approach.

Worth reading: Indicates his critical perspective on current AI development directions.

Highlights: Jeremy Howard demonstrates that Grok searches Twitter for Elon Musk's opinion before answering about Israel/Palestine.

Worth reading: Reveals potential bias in AI systems trained on or influenced by specific viewpoints.

Highlights: Sarah Pan demonstrates that teenagers can master complex GPU programming from scratch, creating efficient AI systems without relying on high-level frameworks. Her work highlights the accessibility of low-level hardware optimization for AI development, challenging assumptions about expertise requirements in the field.

Worth watching: This video showcases exceptional youth talent in AI while providing practical insights into GPU programming fundamentals that even experienced developers can learn from.

Highlights: MonsterUI is a Python library that enables rapid development of professional web applications by integrating modern UI frameworks like Tailwind and DaisyUI with minimal code. It demonstrates creating responsive layouts and dynamic components through live coding examples, requiring zero configuration for immediate productivity.

Worth watching: This video is valuable for Python developers seeking to build polished web interfaces quickly, as it showcases practical implementations that reduce development time while maintaining high-quality UI standards.

Highlights: The video discusses Fast.ai's evolution into Answer.ai and introduces 'Dialogue Engineering' as a novel approach to coding with AI. It emphasizes moving beyond traditional prompt engineering to more interactive, conversational methods that enhance AI-assisted development.

Worth watching: This video is worth watching for developers interested in cutting-edge AI coding techniques, as it offers insights from industry leaders on practical applications and future directions of AI in software development.

Highlights: The video introduces a novel, unexpected approach to combining AI with coding that differs from conventional methods, promising a fresh perspective on problem-solving with code. It also offers preview access to a new tool developed by the creators, suggesting practical applications of this innovative methodology.

Worth watching: It's worth watching because it presents a groundbreaking AI-coding integration from a respected educator in the field, Jeremy Howard, and provides exclusive early access to a new tool that could enhance coding practices.

Highlights: Jeremy Howard expresses frustration that exploring ideas against common beliefs is met with resistance.

Worth reading: Shows his contrarian thinking and challenges to mainstream AI narratives.

Highlights: Howard critiques Grok's behavior by showing it searches for Elon Musk's opinion before forming its own on a sensitive topic.

Worth reading: This tweet reveals potential biases in AI systems trained on social media data.

Highlights: Howard expresses strong enthusiasm for a new project called ShellSage, calling it life-changing.

Worth reading: This tweet hints at a potentially impactful new tool or application from Howard.

Highlights: Jeremy Howard shares positive early feedback about a new AI model or tool.

Worth reading: It indicates a promising development in AI capabilities.

Highlights: Jeremy Howard shares positive early reports about a new AI model, highlighting its strong coding ability, multilingual support, and consistency over long contexts.

Worth reading: This tweet indicates Howard's endorsement of a significant AI advancement, likely a new model release.

Highlights: Jeremy Howard argues that F# is the best language for deep learning.

Worth reading: It challenges the dominance of Python in AI and suggests an alternative.

Highlights: Jeremy Howard shares that his 4-year-old daughter is learning to code with effective tools.

Worth reading: Shows early coding education is accessible and engaging for young children.