Good day, Suwannee County folks! In this edition of our local news, we bring you updates from the world of Artificial Intelligence (AI), a technology that is increasingly becoming part of our everyday lives. Let’s dive in.
A Fireside Chat with Hugging Face’s Thomas Wolf
Yesterday, I had the honor of interviewing Thomas Wolf, the co-founder, and chief scientific officer at Hugging Face, during the CogX Global Leadership Summit in London. For those who don’t know, Hugging Face is the world’s leading repository for open-source AI models—it’s like the GitHub of AI!
Fascinatingly, Wolf emphasized the importance of open-source AI models and how they contribute to the success and impact of AI technology. He believes that smaller AI models, those that can be run on devices like smartphones and tablets, are the future. They offer better privacy, data security, and accessibility, especially for those without constant high-speed broadband connections. Plus, they’re more energy-efficient, reducing AI’s growing carbon footprint and water usage.
Open-Source AI Models: The Key to Democratization
Wolf sees open-source AI and small models as crucial for “democratizing” AI technology. He expressed his concern about how AI has reinforced the power of large technology giants like Microsoft, Google, Amazon, and Meta. He insists that open-source software gives more control to the users, proving vital for businesses to realize a return on investment from AI.
Interestingly, Wolf also touched upon the security risks of open-source AI. He believes that open-source AI will prove to be more secure in the long run—just like how Linux, another open-source software, turned out to be more secure than proprietary software due to the large number of people who can scrutinize and fix the code.
Open Source: A Spectrum
In response to the controversy over Meta’s labeling of its AI software as open source, Wolf suggested that we view openness on a spectrum, with some models being “semi-open”.
Open-Source Models: A Profitable Future?
In an interesting twist, Wolf shared that Hugging Face is cashflow positive, unlike other AI companies. They make money through consulting projects and selling tools for enterprise developers. This hints at a potential profitable future in open-source AI models.
AI in the News
- Geoffrey Hinton and John Hopfield received the Nobel Prize in physics for their contributions to the artificial neural networks that underpin today’s AI revolution.
- Meta introduced Movie Gen, a generative AI model that can create high-quality short videos from text prompts.
- Tim Brooks, a prominent OpenAI researcher, announced his move to Google DeepMind.
- Amazon trained a generative AI model to act as a HR coach for employees.
- OpenAI is considering reducing its dependence on Microsoft’s GPU clusters and has signed a deal with Oracle.
EYE on AI Research
Researchers at the Beijing Academy of Artificial Intelligence showed that simply training a model to predict the next token and training it on multimodal data can produce an AI model that is just as good as those trained in a more complex way. They named their model Emu3.
AI Calendar
Here are some upcoming AI events that might interest you:
- Oct. 22-23: TedAI, San Francisco
- Oct. 28-30: Voice & AI, Arlington, Va.
- Nov. 19-22: Microsoft Ignite, Chicago
- Dec. 2-6: AWS re:Invent, Las Vegas
That’s all for this edition. Stay tuned for more updates from the world of AI!
Jane Sheppard
Note: The facts presented in this article are based on the source provided and do not necessarily reflect the views of Jane Sheppard or her affiliated organizations.
