Ethan Mollick’s new book Co-Intelligence: Living and Working with AI (2024) is hot stuff. After waiting on a long list for a library copy, I finally got my hands on it and breezed through it in a few days. It felt like an extended version of his online presence, and his voice came through consistently and clearly.
If you’re not already following Mollick on LinkedIn, he posts prolifically about AI. Everything from the latest research, his experiments with the tools, how he’s using it in business classrooms in Pennsylvania, how organizations are or aren’t adopting AI effectively, and his musings on where all of this is headed. Somehow he also found time to write a book!
Co-Intelligence is a good starter read for those interested in what Generative AI is all about. It’s also a nice refresher of the technology and Mollick’s top blog posts on the topic. His writing style is accessible and he explains things without getting bogged down in overly technical details that might turn the general public off. This is exactly the kind of approach I believe academics should be taking – distill cutting-edge information into a form that everyone can benefit from, outside of the ivory tower.
My favorite line had to be this: “Humans, walking and talking bags of water and trace chemicals that we are, have managed to convince well-organized sand to pretend to think like us” (p. 193), referring to the silicon that powers the computer and AI age. Regardless of how you feel about Generative AI, this humorously sums up what is happening.
Other favorite parts were when he acknowledged how he had used AI tools to help with the writing process, and showed examples of what the tools had suggested. I appreciated the transparency and the way the tools were presented as helpers to the writer, rather than a replacement. This kind of role-modeling can assist other writers in knowing how to move forward without losing our voice or our skills with the written word.
Mollick even acknowledged the potential in humanities fields (e.g. English, History, Philosophy) for their deep knowledge of cultural history and the potential to create valuable prompts that others wouldn’t think of: “AI could catalyze interest in the humanities as a sought-after field of study, since the knowledge of the humanities makes AI users uniquely qualified to work with the AI.” (p. 116)
Here are some explanations from the book that I liked in their straight-forwardness:
“But in many ways, hallucinations are a deep part of how LLMs work. They don’t store text directly; rather, they store patterns about which tokens are more likely to follow others. That means the AI doesn’t actually ‘know’ anything. It makes up its answers on the fly.” (p. 94)
“If you ask an AI to give you a citation or quote, it is going to generate that quote or citation based on the connections between the data it learned, not retrieve it from memory.” (p. 95)
“This is the paradox of AI creativity. The same feature that makes LLMs unreliable and dangerous for factual work also makes them useful. The real question becomes how to use AI to take advantage of its strengths while avoiding its weaknesses.” (p. 99)
A big part of Mollick’s focus is how AI is changing the nature of work and society. He points out how we take for granted things like organizational structures that are constructs from by-gone eras. For example, the org chart helped run railroads in the 1850s by organizing human workers into clear hierarchies, and its success led to its widespread adoption for bureaucracies in the 20th century (p. 148-149). But it is not the only way of doing things. He discusses ways that AI could radically change how humans work, for better or worse.
He also touches on what he calls the homework apocalypse that has been sweeping education since ChatGPT came out, saying that “AI has come for the king of assignments, the essay” (p. 162). He is hopeful that we will “find a practical consensus that will allow AI to be integrated into the learning process without compromising the development of critical skills” (p. 165).
One thing I didn’t know before is how much assistance he had from his partner, Dr. Lilach Mollick, who he acknowledges developed many of the prompts discussed in the book and gave him key advice. I was pleased to see her role acknowledged. We need visibility of the women involved in the world of AI.
I have recommended this book to colleagues interested in Generative AI as an accessible overview, and look forward to continuing to follow Mollick’s academic adventures engaging with Gen. AI.