Open Source Research and Development

The Open Source Initiative (OSI) has released a new definition for "open-source" AI after a 2 year effort involving consultation with global experts. Founded in 1998, OSI is known for establishing the Open Source Definition, a standard that outlines the criteria software must meet to be considered "open source", e.g. the ability to view, modify, and distribute source code, widely respected in the industry. The new OSI definition for open-source AI requires AI models to make their training data, code, and model weights fully accessible. Definition: "What is Open Source AI When we refer to a “system,” we are speaking both broadly about a fully functional structure and its discrete structural elements. To be considered Open Source, the requirements are the same, whether applied to a system, a model, weights and parameters, or other structural elements. An Open Source AI is an AI system made available under terms and in a way that grant the freedoms to: - Use the system for any purpose and without having to ask for permission. - Study how the system works and inspect its components. - Modify the system for any purpose, including to change its output. - Share the system for others to use with or without modifications, for any purpose. These freedoms apply both to a fully functional system and to discrete elements of a system." The OSI definition also includes specify conditions for open source AI systems, emphasizing the necessity of having access to the preferred form for modifications, which includes: - Data Information: Detailed information about the training data that allows skilled individuals to recreate the system. This includes descriptions of data sources, selection processes, labeling, and methodologies. Public and third-party data sources must also be disclosed. - Code: Full source code used for training and operating the system should be open source, detailing data processing, training procedures, and the architecture of the model. - Parameters: Model parameters, like weights and configurations, should be accessible under open source terms, including details like training checkpoints and final states. There's broad press coverage on this OSI accomplishment, e.g., with these examples mentioned in SiliconANGLE & theCUBE: 1) Meta's Llama Models: These are highlighted as failing to meet the OSI's open-source AI criteria as they have restrictions on commercial use and do not provide open access to the training data or details about it, making it impossible to recreate these models freely. 2) Stability AI's Stable Diffusion Models: Although claimed as "open" by Stability AI, these models require businesses with more than $1 million in annual revenue to purchase an enterprise license. 3) Mistral's Models: Places restrictions on the use of its Mistral 3B and 8B models for certain commercial ventures. On the other hand, these organization have endorsed the new definition: https://lnkd.in/gkeUaQzB

🚀 New Publication Alert! The TODO (OSPO) Group at The Linux Foundation has just published a new paper I authored: “The Lifecycle of an OSPO”, a practical framework to understand how Open Source Program Offices evolve and adapt over time. 📝 Download the paper here: https://lnkd.in/d8R7dkM6 Whether you’re building an OSPO from scratch, scaling one, or reorienting it around new priorities (like AI), this lifecycle model offers a roadmap for structuring strategy, navigating risk, and driving long-term impact. 🔍 The paper explores five key OSPO phases: ✅ Inception ✅ Growth ✅ Maturity ✅ Winding Down ✅ Second Wave (reinvention) This work reflects years of experience working with open source leaders across industries, and I hope it serves as a useful guide for OSPO practitioners, sponsors, and advocates alike. 🛑 The paper also includes sections on the new role of OSPOs as strategic risk radars in light of growing geopolitical tension. I spoke on this topic in previous LinkedIn posts covering how OSPOs in the US, EU, and China are adapting to intensifying geopolitical tensions that affect technology collaboration, data sovereignty, and access to global open source ecosystems. For references, here are the previous posts on this topic for context: ☑️ What US OSPOs must prioritize now? (https://lnkd.in/dDsEpAnw) ☑️ How OSPOs in the EU are stepping up in a shifting regulatory era? (https://lnkd.in/dCzCpBhS) ☑️ Navigating Open Source Amid Geopolitical Complexity | OSPO China Edition (https://lnkd.in/dw2kCsx8) A big thank you to the TODO Group, Ana Jimenez Santamaria 🐧 and TODO's steering committee members for supporting and publishing this paper 🙏 #OpenSource #OSPO #TODOGroup #OpenSourceStrategy The Linux Foundation Linux Foundation Europe Linux Foundation Japan OpenChain Project

Making research reproducible with open-source software During my PhD, I faced a challenge many scientists know well. Results were being published but not reproducible. It made it hard to trust and build on existing results, and I didn't want to be part of that cycle. So, I decided to make my research open and accessible. Enter PyBaMM, an open-source project that grew from this vision. By sharing my code, I aimed to ensure that others could validate and build upon my findings. Here’s why reproducibility matters: - It drives innovation. 🚀 - It builds trust in scientific communities. 📊 - It accelerates progress towards solutions, like achieving net zero. 🌍 At Ionworks, we continue to carry this vision forward. We’re making electrochemical modeling more accessible and reliable, equipping researchers across the globe to tackle some of the biggest challenges in the battery industry. Our open-source approach is about more than just software—it’s about creating a collaborative space where ideas flourish, and solutions are found together. Curious how PyBaMM is reshaping battery science? Follow our work at Ionworks as we push boundaries, bridge gaps, and redefine what transparency and effectiveness look like in science.

In less than two decades, open source software has come to dominate the technology landscape across a wide swathe of key software categories, including operating systems, machine learning, databases, web servers, and more. The open source innovation model has evolved to support a rapidly expanding ecosystem and body of practice supplanting traditional technology development, sales, marketing, and management practices. Open source is also increasingly sparking innovation and forming communities to tackle broad industry problems in diverse fields, including agriculture, public health, motion pictures, and telecommunications. Building on the foundational work of Henry Chesbrough, Eric von Hippel, and Yochai Benkler in Open Innovation, hobbyist technology innovation, and peer-based production, this chapter explores the rise to dominance of open source, the market disruption this emergence has created, and how open source is reshaping legacy business practices not only for early-stage innovation but also for later-stage innovation and collaboration, at scale.

As open source AI accelerates from community movement to real world applications, we are excited to launch our new industry research report on Open Source AI led by Charlotte Xia and Lan W. from my team. This report provides a deep dive into the open source AI ecosystem and its expanding role in AI infrastructure and applications. It examines how open weights are closing the gap with closed APIs, and how this transformation is creating new opportunities in agent infrastructure, security and observability layers, and vertical applications. With OpenAI’s GPT-OSS recent release, the rise of edge-deployable models, and the global race for sovereign AI, open systems are no longer only about transparency - they are becoming the foundation of AI innovation and efficiency deployment. These trends will significantly influence how enterprises build, deploy, and govern AI systems in the years ahead. At Fusion Fund, we continue to study the evolution of open source AI activities and support entrepreneurs building the foundational layers that enable scalable, secure, and efficient deployment. If you would like to connect or discuss new venture ideas, please reach out to Lan W. and Charlotte Xia. Explore the full report in the first comment. #ArtificialIntelligence #OpenSource #AIResearch #AIInnovation #VentureCapital