SeminarTuesday 12 March 2024
6:00pm – 7:00pm
Generative AI and the creation of Large Language Models (LLMs)
Drinks reception: 7:00pm – 8:00pm
13/14 Cornwall Terrace, Outer Circle (entrance facing Regent's Park), London NW1 4QP
Organised by the Daiwa Anglo-Japanese Foundation
Fully bookedThe release of ChatGPT and the broader advent of generative AI was a shocking wake-up call to the world. OpenAI introduced its software in November 2022 and gained over a hundred million registered members within two months. Since then, generative AI has begun to show us the scope of its potential, affecting several industries and changing ways of working, businesses, and ultimately society.
Tech giants like Microsoft and Google are of course in the front line of the competition, expanding large ‘foundation’ models such as LLMs. These models took off particularly around 2014 as increases in available data, more computer resources, and scientific advances in the structures of neural networks created more capable systems. But LLMs do not necessarily need a large number of parameters, and using open sources to develop them has turned out to be even cheaper and faster.
With this in mind, Ito believes that competitive superiority in this field does not lie in the size or sophistication of the model, but in customising it to fit the purposes of the users. What is wanted and is being developed among the most advanced researchers now is the quality of customised models, not the quantity of universal models. This is the future of AI models, which will be safer and more cost-effective.
As a leader in both AI innovation and investment, Ren Ito provided a brief overview of the current state of generative AI, before diving deeper into his own unique insights about the industry. He also provided recommendations about how we might proceed in our innovation and usage of generative AI, introducing notable examples from the UK and Japan and highlighting the importance of international cooperation to deal with existing and emerging challenges and problems.
About the contributor
Ren Ito
Ren Ito is the Co-Founder and Chief Operating Officer of Sakana AI, a Japan-based start-up that draws on understanding about nature-inspired intelligence to deliver powerful AI models at a fraction of the usual cost. Prior to his current role, Ren also served as COO of Stability AI, another AI unicorn based in London.
On top of his experiences directing AI companies, Ren has engaged with the technology industry from an investment perspective in his former capacity as the founder and Partner of Solaris Fund Management. Ren invested in a number of leading unicorn start-ups in the fintech and AI space, including AI unicorns like Afiniti and Stability AI. Additionally, he took the lead on the global expansion of Mercari, Japan’s first unicorn start-up running the fastest-growing marketplace app. He served as CEO of Mercari Europe, as well as successfully orchestrating its IPO, which valued the company at $6bn.
Prior to venturing into entrepreneurship, Ren spent 15 years in diplomacy. He held key positions in the Japanese Foreign Service, where he worked on the security alliance with the US, negotiated a free trade agreement with the EU, and represented Japan on the World Bank board. Ren continues to be involved in public policy debates on technological innovation and international politics as a Senior Fellow at New York University School of Law. Ren is the youngest member of the Trilateral Commission.