In March of 1964, a man named Art Fleming stepped in front of America, and onto the lighted studio set of a new quiz show named Jeopardy!… little did he know, Jeopardy! would go on to become a living pop culture icon. Since then, thousands of episodes have aired, hundreds of thousands of contestants have tried their luck, and history been made… but there’s something else about Jeopardy! that’s of interest to us, and it has to do with artificial intelligence.
See… in the process of becoming a television icon, Jeopardy! inadvertently became something else: a window into the next era of the interplay between man and machine… it foretold this very moment we’re in now. In 2011, an artificial intelligence system developed by IBM, named Watson, competed against human champions… and won. This is a computer versus human beings, in real-time, for the first time. It’s a game, but it’s also not.
IBM has long been at the forefront of new computing technologies… their track record for recognizing what’s coming next speaks for itself. In line with that, they seem to be conjuring a new vision for A.I. that includes more widespread adoption, and it involves a startup called Hugging Face. According to IBM CEO Arvind Krishna, in the future… workers, regular folk, and AI are hand-in-hand, not neck-and-neck.
Hugging Face has been just as busy with funding as with their AI tools, and it shows… the company’s valuation sits at $4.5 billion, with their recent Series D pulling in $235 million. IBM was a significant contributor… that’s because IBM is already hand-in-hand with Hugging Face, as the two are working on a suite of A.I. tools together. Hugging Face handles the natural language processing (NLP) piece.
NLP is a field of computer science that deals with the interaction between computers and human (referred to as “natural”) languages. Hugging Face is an open-source AI platform that makes it easier to build stuff, like chat bots… they’re kind of like a construction set, or library, for NLP models. Just like you can go to a library to borrow a book, you can access NLP models via the Hugging Face platform.
Needless to say… that’s in high demand right now, and we don’t see that changing anytime soon. The company offers libraries of pre-trained NLP models, a collection of NLP datasets, as well as a platform for sharing them, along with NLP models, and free learning resources on NLP… and their products are already being used by businesses, researchers, and developers all over the world.
Well… maybe it’s not just folks all over the world using Hugging Face, seems it’s in the stars too… in the form of something called the Geospatial Foundation Model (GFM). The GFM, the brainchild of a team of researchers from IBM and NASA, is a large language model (LLM) that has been carefully trained on a huge dataset of geospatial information… and it’s free on the Hugging Face website.
Why is that a big deal? Glad you asked… this LLM was trained on over 100 billion words, including text descriptions of satellite images and weather data. This means, because of this partnership, we can use the GFM to identify objects on satellite, predict weather patterns, track the spread of natural disasters, monitor land use, assess crop yields, plan for urban development… basically everything we desperately need right now.
We can envision many more fruitful partnerships in store for Hugging Face, and many more projects that can help us solve our toughest problems. Funding rounds and partnerships involve names like Qualcomm, Nvidia, Intel, AMD, Salesforce, Amazon, Google, AWS, and Sound Ventures. There’s no scenario where we look at this company and don’t see some type of impactful innovation in the tea leaves.
Hugging Face CEO Clément Delangue has been forward about intentions for a Nasdaq listing when the time is right… and they’ll be furiously preparing until then. They already have 500,000 language learning models, 250,000 datasets, and 50,000 organizations on the platform. Semiconductors may have been yesteryear’s generative AI all-stars, but datasets are ready for their day in the sun. Come back next week for more from the IPO space.