Guest AI Lecture and discussion with cognitive scientist John S. Ball
John S. Ball, cognitive scientist, inventor of the brain-based Patom Theory, founder of Pat Inc. (Palo Alto, CA, USA), and author, will give a guest lecture titled »Brain emulation using Patom Theory integrated with linguistic framework, Role and Reference Grammar shows the path forward for the human-machine interface«.
Fr.
28.03.
09.30–11.00
In Kalender speichern
Campus Lichtenberg
Haus 6A Raum 6A.229
Alt-Friedrichsfelde 60
10315 Berlin
Fachrichtung Informatik, FB 2 Duales Studium, und d-cube Institut
Abstract of the Guest Lecture:
Since 2006, John S. Ball has been refining software that recognizes human languages. After a time working on language-to-meaning representation, it became clear that something was lacking and progress was being held back. Then, the missing piece was discovered: the unexploited linguistic framework called Role and Reference Grammar. John S. Ball implemented it on a machine to demonstrate human-like understanding and generation of language. Even though human babies learn their mother tongue, "understanding" by machines remains elusive for even today’s most powerful generative Artificial Intelligence (AI) models. In the guest lecture, John will describe the benefits of an accurate and energy-efficient solution in contrast to today’s Large Language Models (LLMs) that are inaccurate, "hallucinating", with costly ongoing processing.
In his recently published book, "How to Solve AI with Our Brain: The Final Frontier in science", John covers key concepts such as:
- Animal brain evolutionThe human brain extension (language)
- New brain science, Patom Theory
- New linguistic framework, Role and Reference Grammar
- The virtuous circle of implementation based on this cognitive science to provide continuous feedback on brain function. John calls it ‘Deep Symbolics’ because it doesn’t rely on training data.
John S. Ball will cover those and the following topics in his guest lecture:
- What is AI, and what isn’t AI?
- Why is Robotics aligned with animal skills?
- Why are human skills easier than animal skills for a computer?
- LLMs hype is slowing us down
- What is the capacity of Deep Symbolics with respect to language?
- How can we build talking machines in any language in 5 years?
- Demo (if time/practical)
About the speaker
John S. Ball is a cognitive scientist, computer engineer and author. Ever since a collaboration with AI legend Marvin Minsky in the 1980s over his brain science, his specialty area has been applying new and emerging brain-based technologies to computers. At the urging of Minsky, John implemented his brain-based model so machines could comprehend our language.
His corporate IT career was underpinned by his undergrad B.Sc. in Computer Science, Mathematics, Physics and Engineering, and an M.Sc. in Cognitive Science. Starting at IBM, he held national level technical roles managing their largest mainframe computers.
In his new book “How to Solve AI with Our Brain” published in November 2024, John writes about his vision to bring trustworthy language AI to the world.