The Technology, Security & Transparency working group in C20 held a seminar in San Jose, CA, on Friday, June 23, 2023. The topic was “AI: Language Models, Business Implications, Ethical & Legal Transparency”.
The seminar was attended by local technologists with a keen interest in the topics of discussion and wanting to identify how they could contribute to the working group’s initiatives.
Swami Dayamritananda Puri, C20 Working Groups Advisor and Executive Director of MA Centers, offered the introductory comments about G20/C20, its working groups and the significance of having Amma as its Chair this year.
Alison Richards, the International Advisor for the TST working group and Director of oneAPI Initiative at Intel Corporation, followed with an update on the efforts and accomplishments of the working group thus far. She introduced the working group’s initiatives GLIDES & TACTICS to the attendees.
There were three main speakers who spoke on the seminar discussion topics.
Dr. Manish Gupta kicked off the discussion with a talk on “Large Language Models Demystified”, where he described the evolution of the language models being used in AI, its benefits, shortcomings and risks.
Manish is currently a staff Research Scientist at Google leading research endeavors in AI, particularly Natural Language Processing and Recommender Systems. Previously, Dr. Gupta worked at Amazon (Alexa) leading several projects in the areas of Conversational AI and Information Retrieval. Dr. Gupta has a Ph.D. in Applied Mathematics from Harvard University and B.Tech from IIT Madras.
Mr. Krithik Puthalath talked about “Inevitability of AI in our daily lives and how it can be advantageous”. Krithik highlighted that the AI revolution is still in its infancy with a lot more work to be done. He also highlighted the infrastructure costs associated with AI models and the need for optimizing it.
Krithik is the Founder & CEO of a startup called Zyphra Technologies. The company builds software, systems and next generation microprocessors for AI. Zyphra’s solution has the potential to reduce the cost of training AI models by orders of magnitude, democratizing AI to leading enterprises and governments. Krithik has a Bachelor’s degree in Materials Science and Electrical & Electronic Engineering and a Master’s degree in Theoretical and Mathematical Physics from the University of Illinois Urbana – Champaign.
Ms. Nisha Ramachandran concluded the series with her perspective on “Transparency & AI and how it impacts business”. She discussed why it’s a little difficult for companies to create frameworks to be followed so as to ensure that transparency is maintained in their work when using and/or creating AI models, since there are no existing precedents. She also indicated that it is one of the reasons why governments across the world are developing regulations to manage the use of this technology.
Nisha is the General Counsel at Signifyd, a company which offers commerce protection that recognizes the identity and intent behind every payment, restoring mutual trust between merchants and their customers. Prior to Signifyd, Nisha was the Associate General Counsel at Fitbit, which became a part of Google.
Nisha completed her Bachelor’s degree in History & Cognitive Science, Master’s degree in Public Policy and Doctorate of Law degree from the University of California, Berkeley.
Alison Richards delivered the closing comments and encouraged the attendees to get involved in the initiatives with their thoughts and feedback.