Video and slides for talk on human-like AI

14 June 2023 talk to the Darmstadt Ontology Group.

Human-like AI: from logic to argumentation, reasoning with imperfect knowledge in the era of AGI

Human interactions and our understanding of the world is replete with uncertainty, imprecision, incomplete and inconsistent knowledge. Logicians have largely turned a blind eye to the challenges of imperfect knowledge. This is despite a long tradition of work on argumentation, stretching all the way back to Ancient Greece, that underpins courtroom proceedings, ethical guidelines, political discussion and everyday arguments.  I will introduce the plausible knowledge notation as a potential solution that covers plausible inferences of properties and relationships, fuzzy scalars and quantifiers, along with analogical reasoning. I will then demonstrate that large language models are good at plausible reasoning, and examine the role of symbolic knowledge in the dawning era of artificial general intelligence. Computers are already better than us at dealing with complex knowledge on a large scale. We can look forward to collaborative development of knowledge based systems using AGIs working hand in hand with human experts.  Today’s large language models are just the beginning of an intellectual revolution that will have huge and disruptive effects on work, education and play.

Video: https://youtu.be/GMDGbRQSWyo
Slides: http://www.w3.org/2023/06-Human-like-AI.pdf

p.s. I am behind in copying recent work to the GitHub repository and hope to get that done in the next few weeks as I progress the work on cognitive swarms. I also want to draft a Community Group report on PKN - the plausible knowledge notation similar to the existing report on Chunks & Rules.

Best regards to all.

Dave Raggett <dsr@w3.org>