The TAICO team is proud to announce our next meetup on Wednesday, November 27th at the Adaptavist office in Toronto. Much thanks to Adaptavist for hosting!
đź”— tldr - event registration link - meetup.com
Registration and Event Details
Please note that we are using Meetup.com for event registration and you must register through that link to attend. Seating is limited so please register early. Thank you and see you there!
Event registration link - meetup.com
Our goal is to bring artificial intelligence and cybersecurity together. To do that, we need to explore what’s happening in AI and what’s happening in cybersecurity, and where the two intersect and collide. We’re also working to understand how we solve problems in these areas, what that looks like, and how Canada can and will participate.
With that in mind, we’re pleased to announce our speaker and the agenda for our next meeting!
Agenda
- đź‘‹ Welcome and introductions
- 🚀 Demo of GenAI and security in action
🎤 Speaker #1: David Sampson CISSP
Talk Title: Application Security and Artificial Intelligence
About David:
“David Sampson is the Director of Information Security at Globys, where he leads the company’s cybersecurity strategy and runs its cutting-edge AI initiative. With a background as an Application Security Engineer, David is passionate about integrating security into the development lifecycle, leveraging AI tools like ChatGPT and GitHub Copilot to enhance secure coding practices. His leadership was instrumental in achieving Globys’ ISO 27001 certification, and he continues to innovate in both security and AI fields.
In addition to his role at Globys, David is the founder of Perdition Security, a firm dedicated to providing advanced cybersecurity services to small and medium businesses. He is a strong advocate for developer enablement, focusing on practical ways to embed security into everyday workflows. As a frequent speaker at industry conferences, David shares his insights on AI, cybersecurity, and the future of secure development.
Abstract:
“In this session, I will explore how security teams can leverage the enthusiasm for generative AI to teach developers how to integrate robust security measures into their code seamlessly. Focusing on the DevSecOps cycle, I’ll demonstrate how generative AI can assist in key security activities such as finding vulnerabilities, explaining security flaws, and providing actionable fixes. Using real-world technical examples, I will walk through how AI-driven tools can enhance the efficiency and accuracy of security practices in development environments, making security an accessible and integral part of the development lifecycle.”
🎤 Speaker #2: Joshua Carpeggiani
Talk Title: LLM Hallucinations; What are they and how can they be mitigated?
Abstract:
“This talk presents a comprehensive overview of hallucinations in Large Language Models (LLMs), one of the most critical challenges facing modern AI systems. We begin by introducing a refined taxonomy of hallucinations, categorizing them into factuality hallucinations (inconsistencies with real-world facts) and faithfulness hallucinations (deviations from given contexts or instructions).
We then explore the multifaceted causes of hallucinations, from data-related issues like flawed training sources and knowledge boundaries, to training-related challenges in both pre-training and alignment stages, to inference-time complications arising from decoding strategies.
The presentation will cover state-of-the-art detection methods and evaluation benchmarks, examining how researchers are working to identify and measure hallucinations in LLM outputs. We’ll discuss various mitigation strategies, including retrieval augmentation, knowledge editing, and enhanced decoding techniques. Finally, we’ll address emerging challenges in specific domains like long-form text generation and multimodal systems, and explore open questions about LLMs’ self-correction capabilities and the balance between creativity and factuality. This talk aims to provide researchers and practitioners with a comprehensive understanding of this crucial challenge in AI development.”
Talk References:
- https://arxiv.org/pdf/2311.05232
- https://arxiv.org/abs/2406.02543
- https://arxiv.org/pdf/2307.10236
- ⚡ Lightning talks and demos - Please let us know if you have anything to show
- 🍻 Networking and drinks
⚡ Lightning Talks and Demos
Please reach out to us if you’d like to present at the meetup. We are looking for people to talk about what they are working on, what they are building and learning, and are open to any level of experience and technical depth. Whether you are a beginner or an expert, we want to hear from you! We’re all just out here building and learning.
đź‘‹ See You There!
Thanks, and we look forward to seeing you at the meetup!
