Talks on using AI reliably.
I help non-technical audiences use AI reliably: students, teachers, librarians, communicators, and public servants who need answers they can trust.
AI answers with the same confidence whether it's right or wrong.
AI tools are now part of how people study, write, and make decisions. They invent sources, state errors as fact, and repeat outdated information. Most people have never been shown how to check.
People are going to use these tools regardless. What they need is a quick, dependable way to tell a good answer from a bad one, before a wrong one ends up in an essay, a news story, or a public decision.
How to get the best answer from AI.
Most people ask an AI tool a question once and take what they get. This talk shows a better habit. Ask the same question five ways, compare the answers, and you can see how far to trust the result. The method takes ten minutes, needs no technical skill, and comes from published research testing four major AI tools on real documents.
Audiences leave with
- A five-step method they can use the same day
- A clear sense of where AI tools go wrong: invented sources, confident errors, and outdated information
- A free printed or digital guide to keep and share
Formats & fees.
- Short talk20 minutes plus questions, virtual
- Guest lecture or keynote45 to 60 minutes, virtual or in person
- Hands-on workshop90 minutes, virtual or in person
Short talks start at $500. Longer sessions and workshops are quoted by format and audience, and in-person events add travel costs. I regularly work with classes, libraries, and community groups, and I offer reduced rates for them, so please ask even if your budget is small.
Built for general audiences.
- University and high school classes
- Teacher training and professional development days
- Library staff and library associations
- Communications, journalism, and public-sector teams
- Conferences that want a practical AI session for a general audience
An independent researcher who tests the tools.
Katryna Peart is an independent researcher and writer who tests how AI tools handle history, government documents, and public information. She has evaluated AI systems for Google, Uber AI, and Mercor, and her research has been accepted at the BCS SGAI International Conference on AI. She writes for Governing, Route Fifty, PM Magazine, and PublicTechnology, and holds a master's degree in history from Royal Holloway, University of London.
Talks and appearances in 2026
- Guest lecture, Columbia University (October 2026)"How to Get the Best Answer from AI," for the AI, Information, and Meaning module
- STARWEST 2026
- PublicTechnology PT Exchange (October 2026)
- Ars Inquirendi, St Edmund Hall, Oxford (November 22, 2026)"The Digital Priesthood: Archival Silence and the Cumulative Compression of Pre-Print Knowledge"
- BCS SGAI International Conference on AI, Cambridge, UK (December 2026)
Book a talk.
Write with your date, audience, and format, and I'll get back to you. I also test AI tools for organizations, so ask if you would like yours checked.