AI Augmentation vs. Automation
An Experiment with NotebookLM
November 30, 2024, Vol. 2, Issue 50
Cite as: Coleman, Anita S. (2024, November 30). AI Augmentation vs. Automation: An Experiment with NotebookLM. Infophilia, a positive psychology of information, 2 (50).
✨ Welcome, new and returning readers, to Infophilia! I’m currently battling the flu (a side effect of a month of travel), but I’m grateful for a peaceful Thanksgiving weekend full of warm sunshine and family. I hope this finds you in good health and good cheer.
Before diving into today’s feature, I share this link to a quick Note on Librarians / Library Voices on Substack that I wrote 11/21/24. Some of us on one of the American Library Association Connect online discussion forums had been discussing alternative forms of media, scholarly communication, and publishing and I thought this list may be helpful. It's not yet comprehensive, but it’s a starting point.
That exploration also got me into my recent experiment with NotebookLM, a tool that I’ve admittedly neglected in my own research until now. When I uploaded my top ten Infophilia Substack articles to NotebookLM, I had a general sense of its capabilities as a research assistant. I wasn’t so much drawn to the hype about its ability to help people learn faster or generate polished writing from notes, but rather to its potential to make connections across diverse sources. I may have unintentionally introduced some bias into my experiment by feeding it the top ten most downloaded essays from Infophilia, which might have skewed the results. Still, what I didn’t anticipate was the stark demonstration of both the impressive capabilities and the significant limitations of generative artificial intelligence (Gen AI). Within minutes, the Google product had generated an 11-minute podcast, a comprehensive study guide, and multiple supplementary documents - FAQ, table of contents, timeline, and a briefing document. It was impressive—and unsettling.
Augmentation vs. Automation in Generative AI
While NotebookLM promises to augment human intellect by providing personalized assistance in research, it’s important to differentiate between augmentation and automation—two key capabilities of AI. Augmentation refers to AI’s ability to enhance and extend human cognitive abilities, offering insights, suggestions, and tools that empower us to process information and work more effectively and creatively. In contrast, automation involves AI taking over specific tasks or processes entirely, often without human input or intervention. For general readers, understanding this distinction is crucial because it shapes how we can engage with AI in everyday life—from simplifying tasks to transforming how we think, read, work, and make decisions.
The strength of tools like NotebookLM lies in their potential for augmentation: helping us process information, generate new ideas, and deepen our understanding. However, the challenge arises when AI blurs these lines, sometimes crossing over into automation, where it reduces complex tasks to simple outputs that lack the nuance and depth of human engagement. Understanding the balance between these two capabilities is essential for evaluating the true value AI brings to intellectual work.
In my case, NotebookLM excelled at augmenting my work by providing an audio overview, summaries in many different forms, and generating thought-provoking questions, but it also veered into automation. As I reviewed the AI-generated content, I realized I was witnessing something profound: a powerful assistant that could simultaneously illuminate and flatten the intellectual landscape of my work.