Nathan, Joelma, and Andrei Torres (adjunct prof. Ontario Tech) delivered a public workshop at Ajax Public Library this week.
The workshop explored the limitations, risks, and vulnerabilities of large language models (LLMs), some defenses, and, how such defenses are often insufficient. We demonstrated this through some hands-on activities and by taking down Gandalf in the AI Wizard game.
The hands-on workshop focused on how AI can be impressive, useless, and/or weird via exploration, experimentation, and co-creation of knowledge. In doing so, attendees could discover the limits of:
This event marks one of many library events this year, you can find out more here.
On Tuesday, April 14, the Trustworthy AI lab attended the MAIRI symposium that featured inspiring speakers, a poster session, a panel discussion, and opportunities to connect with a community committed to thoughtful AI research and innovation. Among the various talks, many in the lab presented their research in the poster sessions.
Last week, Peter gave the inaugural talk for the Autonomic Computing and Self-Organization Systems (ACSOS) Seminar Series, hosted by the Translational Data Analytics Institute on The Ohio State University Campus.
Peter’s event delivered an engaging talk on Reflective Social Intelligence for Autonomous Agents. The session highlighted the importance of explicitly modeling social phenomena, concepts, and emergent dynamics within autonomous systems, such that said systems may become aware and act intentionally because of said knowledge.
A huge congratulations to Dr. Joelma Peixoto on winning the Postdoctoral Fellow Excellence Award!
The School of Graduate and Postdoctoral Studies (SGPS) established the Postdoctoral Fellow Excellence Award in 2019 to recognize postdoctoral fellows who have displayed excellence in research, professional development, and/or service during their appointment at Ontario Tech University.