AICoP: Google - Beyond LLMs in Higher Education

Columbia University’s AI Community of Practice Explores Google’s Latest AI Innovations

The AI Community of Practice at Columbia University, led by the Emerging Technologies team, convened for its February 2025 session, featuring an in-depth presentation from Google on the evolving role of artificial intelligence in education, research, and enterprise applications.

The session, hosted by Charles Elliott, a Google AI expert, provided attendees with insights into the latest Gemini AI models, Multimodal capabilities, Agentic AI, and AI-powered research tools. With AI continuing to transform higher education, the discussion centered on how Google’s cutting-edge technologies can enhance productivity, decision-making, and academic workflows.

Google’s AI in Higher Education

Elliott introduced Notebook LM, Google’s AI-powered note-taking and document analysis tool, designed to help students and researchers manage complex information with greater accuracy. By keeping AI responses grounded in uploaded documents, Notebook LM minimizes misinformation and enhances study guide generation, interactive Q&A, and AI-driven audio summaries.

The session also delved into Google’s advancements in Agentic AI, a next-generation approach that enables AI systems to autonomously perform complex tasks. Potential applications in education include:

• AI-assisted curriculum design, where AI tailors coursework based on student learning patterns.
• AI-powered administrative assistants for streamlining university processes like document handling and transcript processing.
• AI-driven research assistants, capable of generating and ranking hypotheses to support scientific discovery.

Live Demonstrations and Key Innovations

Attendees saw firsthand how Google’s latest AI models push the boundaries of text, image, and video understanding. Highlights included:

• Multimodal AI capabilities, with Google Maps AI integration for real-time geospatial insights and historical analysis.
• Advanced image processing, demonstrating how AI can edit photos, recognize visual patterns, and enhance digital content creation.
• AI-generated videos, showcasing how Google’s V02 model is redefining creative applications in media and research.

The presentation also touched on Google’s Co-Scientist initiative, which enables AI models to assist researchers in forming and testing new scientific hypotheses.

Addressing AI Challenges and Ethical Considerations

The session tackled pressing concerns in AI adoption, including:

• Data Privacy & Compliance – Google reaffirmed its policy of not using customer data for training in enterprise and educational applications.
• AI Transparency & Accuracy – Strategies were discussed to ensure AI-generated responses remain grounded in verifiable sources.
• Institutional AI Integration – Universities must navigate how to deploy AI responsibly while ensuring compliance with security protocols.

The Future of AI at Columbia

Looking ahead, Columbia faculty and students are encouraged to explore Google’s AI tools and integrate them into research, teaching, and administrative workflows. The AI Community of Practice will continue its monthly sessions, fostering ongoing discussions about the evolving AI landscape.

For those interested in testing Google’s AI solutions or joining upcoming discussions, the Emerging Technologies team can be reached at [email protected].

Presentation deck can be found here.