A poem in English is translated into another language…the translated version maintains the original concept and rhythm using words found in the language into which it was translated. A machine completed this translation in seconds. How can this be?
Dr Peter DeJong will explain for us the use of AI and the intellectual performance of Large Language Models which produce such results. The results are surprising and unexpected even to those researching and developing the AI world.
In the last 10 years, computers have surpassed humans in the performance of many tasks that previously required human intelligence to master; this is the result of advances in machine learning technology. Machine learning is a branch of computer science which uses data and algorithms to imitate how how humans learn. Dr DeJong will give an overview of how these technologies work, and the relationship of machine learning to human cognition.
Dr deJong has a PhD from MIT in Artificial Intelligence with a minor in Cognitive Science. In the 1960s he architected, managed, and built large systems for Boeing, IBM, and Shell Oil. In the 1970s, he was a research scientist at IBM Research. There he started project which created IBM’s first Relational Data Base System. In the 1980s, he was a PhD student in MIT’s Artificial Intelligence Laboratory. At MIT, his research focused on highly parallel and distributed AI reasoning systems. After obtaining his PHD he worked on parallel and distributed systems for IBM, Hewlett-Packard, and Microsoft.
Co-sponsored by Bethesda Metro Area village and the Connie Morella Library. Free and open to all. Register here for a Zoom link. https://mcpl.libnet.info/event/9630550