Summer Festival of AI

9:00 SA - 10:00 SA (UTC)

This session introduces the foundational elements of prompting, helping participants understand how to communicate effectively with AI. It covers essential technical concepts such as context windows, cutoff dates, hallucinations, and parameters, offering a clear understanding of how these factors impact outputs. Participants will explore the background of Generative AI and learn best practices for writing prompts that are clear, goal-oriented, and structured for maximum effectiveness. The CREATE model is introduced as a framework for prompt construction, along with the use of personas to enrich interactions. Key tips for working with tools like Copilot are shared, and common prompting mistakes are discussed to help learners avoid them. Attendees are encouraged to treat AI like a smart intern—useful, capable, and guided by human judgment. The session emphasizes three key stages of prompting: preparation, delivery, and refinement, reinforced by the simple and effective 555 Rule. Finally, it addresses the ethical use of AI and how to stay safe when engaging with these tools.

9:00 SA - 10:00 SA (UTC)

Building on the foundations of the 101 session, this intermediate session deepens participants’ understanding by introducing reasoning models that support more sophisticated queries. Learners are guided through the use of Deep Research models for complex investigations and are taught how to use example-based prompting (shots) to achieve better output. The importance of follow-up questions is emphasised to iteratively refine responses. Speech models are introduced, enabling voice interaction with AI, and participants gain hands-on experience with Google Notebook LM and interactive podcasts. Editing capabilities through Canvas modes are also covered. Attendees will develop new prompting styles and learn how to build writing consistency and tone. The session encourages collaboration through tools like Spaces and sharing features, while also offering practice tools such as Google's "Say what you see" technique. Advanced prompting practices—like iteration, handling conflicting personas, and applying self-critique—are explored. The session concludes with a focused look at emerging risks and how to safely navigate them.