• AI Time to Impact
  • Posts
  • . . AI: Hugging Face Reverts to Apache 2.0 and AI Investment in the UK and Canada (4.8.24)

. . AI: Hugging Face Reverts to Apache 2.0 and AI Investment in the UK and Canada (4.8.24)

Google, Hugging Face, UK, Microsoft AI, Canada

Author

Marshall Kirkpatrick
April 08, 2024

Friends, today's news includes a tour around the world in major AI investments, license changes, and technology upgrades. 

We're trying a new format, where only the 2 or 3 most important and interesting stories get full summaries, then we'll just link one-liners to the other important news of the day.  We hope this will be easier to read and all the more scalable to write.

As always, these are the AI stories that the AI community is talking about the most. We hope you’ll find some of them directly and immediately useful, others as sources of inspiration that will unfold over time in the future.

Here’s today’s news.

-Marshall Kirkpatrick, Editor

Hugging Face Reverts TGI to Apache 2.0 Open-Source License

First impacted: Software Developers, Open-Source Community Contributors
Time to impact: Short

Hugging Face has announced a switch in the license for its popular toolkit Text Generation Inference (which enables text generation for LLMs from Llama to Falcon to StarCoder and more) from a custom commercial license called HFOIL back to Apache 2.0, making the software library fully open-source once again. According to a series of tweets by CTO Julien Chaumond, this decision was made after a trial with a custom license, initiated in July 2023, failed to generate new business opportunities and complicated community contributions due to legal uncertainties. [via @julien_c] Explore more of our coverage of: Open-Source Software, Text-Generation-Inference, Apache 2 License. Share this story by email

First impacted: AI scientists, U.K. workforce
Time to impact: Medium

Microsoft AI has launched a new AI hub in London, led by former DeepMind researcher and AI scientist Jordan Hoffmann, focusing on advancing language models and their supporting infrastructure. This move is part of Microsoft's commitment to invest £2.5 billion in the U.K. to prepare its workforce for the AI era and to bring 20,000 top-tier GPUs to the country by 2026. [Announcing new Microsoft AI Hub in London - The Official Microsoft Blog] Explore more of our coverage of: Microsoft AI, Language Models, UK Tech Investment. Share this story by email

Explore more of our coverage of: Canada AI Investment, AI Infrastructure, AI Safety Institute. Share this story by email

Explore more of our coverage of: AI Model Update, RL Algorithm Change, Gemma IT Chat. Share this story by email

Explore more of our coverage of: PyTorch Training, Schedule-Free Learning, AI Optimization Techniques. Share this story by email