


GenAI models not always factual, issue needs to be worked on, fixed: Google’s Jeff Dean
Jeff Dean, chief scientist, Google DeepMind and Google Research, spearheads Alphabet Inc’s recently merged artificial intelligence groups – DeepMind and Google Brain. Among its most famous engineers, Dean joined Google in 1999 and has since been at the forefront of AI research in Silicon Valley. Along with Google DeepMind CEO Demis Hassabis, he now leads all of Google’s critical and strategic AI projects, as outlined by Alphabet’s CEO Sundar Pichai earlier in April.
In an interview with ET’s Samidha Sharma at the company’s headquarters in Mountain View, California, he spoke about playing catch-up in the artificial general intelligence (AGI) race, its chatbot Bard's launch, impending release of Google’s Gemini — a ChatGPT competitor — as well as strategies to combat misinformation fuelled by AI tools. Edited Excerpts.
Elevate Your Tech Prowess with High-Value Skill Courses
Offering CollegeCourseWebsiteIIM LucknowIIML Executive Programme in FinTech, Banking & Applied Risk ManagementVisitIndian School of BusinessISB Professional Certificate in Product ManagementVisitIIM KozhikodeIIMK Advanced Data Science For ManagersVisitYou have seen AI evolve over many years. What’s your view on the exuberance around Generative AI?
A lot of these generative AI models are built on research that was conducted at Google since 2016, on top of the model architecture called The Transformer. This model enabled very efficient learning from sequences compared to previous approaches. What has happened in the last year is that a few different companies have put out interfaces, where end consumers can play with these models, and it’s become far more useful for a broader set of people.
Google was not as quick to launch products in the Generative