One critical debate is whether the future of GenAI models is primarily open source. This ties into other debates like whether there are diminishing returns to continued increases in model size or the efficacy of large commercial models versus smaller, more targeted models.
No opinions to offer beyond noting these developments merit tracking.
Recommended Reading
The emirate of Abu Dhabi is making a large-scale artificial intelligence model, "Falcon 40B", available open source for research and commercial use, the government's Advanced Technology Research Council (ATRC) said on Thursday.
Meta is continuing to deliver high-quality research artifacts and not backing down from pressure against open source.
What is Open Source in AI? Metas LLaMa 2? No, says the Open Source Initiative.
Meta Platforms is preparing to launch software to help developers automatically generate programming code, a challenge to proprietary software from OpenAI, Google and others, according to two people with direct knowledge of the product. Meta’s code-generating artificial intelligence model, ...
/PRNewswire/ -- Today, SAP SE (NYSE: SAP) and Google Cloud announced an extensive expansion of their partnership, introducing a comprehensive open data...
Cloud platform startup Together AI has raised $102.5 million in a Series A funding round led by Kleiner Perkins. The company helps companies pre-train and fine-tune open source foundation models more efficiently on expensive graphics chips. Together AI's team includes experienced AI researchers and its chief scientist, Tri Dao, created an open-source library called FlashAttention. The startup aims to be a key player in the open source resistance to big, closed foundation model companies like OpenAI and Anthropic.
Mistral AI has made its first large language model free for everyone. The model was released on December 25, 2023, and the article provides more details. Article Link [https://techcrunch.com/2023/09/27/mistral-ai-makes-its-first-large-language-model-free-for-everyone/]
A new open-source language model called "Smaug-72B" has claimed the title of the best in the world, outperforming GPT-3.5 and Mistral Medium in popular benchmarks. Developed by Abacus AI, Smaug-72B is a fine-tuned version of Qwen-72B and excels in reasoning and math tasks. This achievement signals the potential of open-source AI to rival Big Tech's capabilities and reshape the AI landscape. Qwen has also released Qwen 1.5, a suite of powerful language models. The emergence of these models has sparked excitement and debate, highlighting the rapid evolution of open-source AI and its potential to challenge tech giants.
Corporate open source is facing challenges as companies like HashiCorp and Redis move away from open source licenses. The article discusses the impact of these changes on the open source community and the need for trust and transparency. It also highlights the opportunity for new open source projects to emerge without compromising on freedom and philosophy. The author suggests avoiding Contributor License Agreements (CLAs) and sticking to open source licenses that respect user freedom.
Meta released their latest open-source language model, Llama 3, which is considered one of the best models available. The article discusses Meta's strategy of investing in open-source AI to control their value chain and prevent new entrants from gaining disproportionate control. By commoditizing the model layer and making it easily accessible and free, Meta aims to attract developers and drive a better ecosystem. This approach aligns with Meta's application-focused business model and helps them avoid reliance on a single model provider.
The article discusses the recent surge in open-source large language models (LLMs) in the field of AI. It highlights specific launches of open-source LLMs by companies like Databricks, A21 Labs, Alibaba Cloud, SambaNova Systems, NexusFlow, xAI, Mistral, and Mobius Labs. The author emphasizes the significance of these launches in the diversification and proliferation of accessible and decentralized AI models. The article also mentions the narrowing gap between open-source and closed-source AI technologies.
The tech industry is struggling to define "open-source AI," leading to confusion and potential misuse of the term. The Open Source Initiative (OSI) is working on a definition, but the complexity of AI models and the lack of agreement on criteria pose challenges. Companies like Meta, OpenAI, and Google have differing approaches to openness, with some releasing pretrained models but not the training data. The debate revolves around access to data, restrictions on use, and the need for transparency. A clear definition is crucial for fostering innovation, competition, and user freedoms in the AI community.
Previous
2 / 2