Book Review: Co-Intelligence (Ethan Mollick)
As alien as AIs are, they’re also deeply human. They are training on our cultural history, and reinforcement learning from humans aligns them to our goals. They carry our biases and are created out of a complex mix of idealism, entrepreneurial spirit, and yes, exploitation of the work and labor of others. In many ways, the magic of AIs is that they can convince us, even knowing better, that we are in some way talking to another mind. (pg. 211)
In this brilliant, highly accessible and leading review of Generative Artificial Intelligence (GAI), Ethan Mollick bring the reader from a relative novice understanding of GAI, into the weeds of how they are currently, and might, shape the way we work, and our future. I can highly recommend this book for a number of reasons, but would consider this required material for the 2024-2025 academic year (that is, until it is updated!)
You would be interested in this book if you…
(1) Have or want to experiment with GAI in your personal and professional practice
(2) Want to have some excellent provocations for engaging in the discussion about the place and future of GAI in your own life and in education writ large
(3) Wanted to expand your thinking on how, when and where to engage with GAI
What is Co-Intelligence?
“Workers who figure out how to make AI useful for their jobs will have a large impact…Humans are subject to all sorts of biases that impact our decision-making. But many of these biasses comr from our being stuck in our own minds. Now we have another (strange, artificial) co-intelligence we can turn to for help” (pg. 48)
Mollick argues for GAI, such as ChatGPT and other platforms, to take a place alongside us. Doing this effectively will allow us to expand our capacity and impact in our jobs. To this he also offers 4 Rules for Co-Intelligence:
(1) Invite GAI to everything: while there is no user manual, we can learn how to best utilize GAI by trying it out for anything and everything. As we use it more and more, we will become more discerning for how and when to use it. But for now, invite GAI to all of your tasks.
(2) Be the Human in the Loop: inviting GAI to everything doesn’t mean letting it take the wheel. GAI has an operative principle: “It can help to think of the AI as trying to…”make you happy” by providing an answer that you will like.” (pg. 52) This is one reason why AIs are known to hallucinate and make up answer. They want to give you what you want. Being the human in the loop requires constructive doubt, critical thinking and ethical considerations. (pg. 54)
(3) Treat AI like a Person: Do this not because it is a person, but because it is drawing from very human data, trillions of data points that are generated by and for humans. When you are prompting GAI to do something, taking on a relationship, conversational feel, is getting the best results in these early renditions. Give it a persona, and Mollick argues that GAIs even respond to emotional manipulation. GAIs, “produce better answers if you tell them ‘this is important to my career’ as part of your prompt.” (pg. 57)
(4) Assume this is the worst AI you will ever use: As you invite GAI to all the things you are trying to accomplish, and while you are being the human-in-the-loop, it is helpful tokeep in mind that GAIs will be getting better and better with each upgrade. So hone these skills, but you should also be considering where the GAI capabilities will be going in the near future.
The Future of Work
I think that this is one of the strongest, most compelling reasons to read this book. Throughout the book, Mollick hints at what the future of the world of work might be. Everything from GAIs as accelerative creativity and innovation, to a world of disinformation and economic costs and benefits to increased efficiency.
AI as a Creative is an entire section of the book, where he explains that GAIs are connection-machines. Connecting disparate ideas in meaningful ways is a key foundation to creation and innovation. In a human-judged competition, “…the degree of victory was startling: 40 best ideas rated by the [human] judges, 35 came from ChatGPT.” (pg. 105) This, Mollick argues, could reinvigorate creativity and innovation in a world that is slowing down in both of these fields.
In a world where meaning is getting more and more diluted, how might GAI play a role? Mollick argues that GAI can play a role in exposing the facade of some “meaningful tasks” that really are meaningless. For example, and this may be contentious, the REPORT CARD COMMENT. Research shows that a very small percentage of families and students read report card comments, and a decreasing amount of educators are putting personal meaning into them. So can GAI take on this task? Should it? These are two very different questions that are being provoked by GAI in education.
Meaning is being shaped by the sheer volume of information that is being produced on the world wide web. Mollick addresses this throughout the book, writing about the role that GAI plays in amplifying disinformation by literally writing the data that it is drawing from. For more on this check out the MIT Technology Review article: “How generative AI is boosting the spread of disinformation and propaganda” where you can read more about how, “In a new report released by Freedom House, a human rights advocacy group, researchers documented the use of generative AI in 16 countries “to sow doubt, smear opponents, or influence public debate.”
Mollick addresses the possible social and economic impacts of GAI. “The adjustment to this shift [of GAI taking on tasks throughout society] if it were to occur, is hard to imagine. It will require a major rethinking of how we approach work and society. Shortened weeks, universal basic income, and other policy changes might become a reality as the need for human work decreases of over time. We will need to find new ways to occupy our free time in meaningful ways, since so much of our current life is focused around work.” (pg. 206)
As for what the future may look like, Mollick paints three different scenarios of possible futures, where GAI evolves no further than what it is currently; where GAI undergoes linear rate of evolution; and, where GAI underdoes exponential rate of evolution. Finally, he projects a future where GAI evolves into The Machine God. Each of these scenarios is worth reading, because we, as humans-in-the-loop, play a critical role. We can’t leave this up to governments and bad actors.
The Future of Education: Our Greatest Hope
If you’ve read Sal Khan’s “Brave New Words”, or read my review of it, some of what the author discusses will be familiar; however, what sets this book apart is the lived experience of Mollick working with, and getting real-time feedback from his post-secondary students.
He argues that schools have an ever-increasing race against the build-up of content that the nature of education must change. He argues that we need to build expertise, engage is deliberate practice, and receive feedback on that practice. This pathway requires a grounding in facts (pg. 181) and a connection between these facts (pg. 182), and this is most effectively done with teachers that know their students. Where GAI can help, is in providing students with the content through a GAI flipped classroom, where the teacher then supports the students in making the connections and creatively bringing the content to life. The teacher then sets up deliberate practice and application of the content into novel situations, where, again, the GAI provides real-time feedback. (pg. 183)
If field after field, we are finding that a human working with an AI co-intelligence outperforms all but the best humans working without an AI. (pg. 189)
This book will pull out the most stubborn of heads from the sand. It is not a dooms-day book about Generative AI. Rather, it is a thoughtful exploration with real (and sometimes hilarious) examples of how GAI was Mollick’s own co-intelligence in the writing of this book. At one point, when researching articles he himself had written, ChatGPT was responded with article so well faked that Mollick had to go through his own old files to verify that he hadn’t written said articles. He also writes with hope about a future with GAI for our students and for our societies. And within this is a call for action – to learn and experiment with GAI in our own lives and roles. This is a call to action that I am taking. Will you?
