When I start working with a school or an organization in the area of artificial intelligence, it’s usually because they say that they want to use it to become more effective or efficient. They want to save time and money and find a way to alleviate some of the workload on the humans.
Or, sometimes, it’s from a place of curiosity. They’ve seen stories or examples of cool (or weird) things that AI can do and want to learn more. They don’t totally understand what AI is and want a basic foundation.
Really, though, the question on everyone’s mind is, “Will I be replaced? And how quickly?”
They worry about becoming obsolete.
This is especially relevant for my clients because for most educators, what they do is more than a paycheck; it’s a calling and a passion and a core part of their identity.
Enter co-intelligence.
Ethan Mollicks’s new book Co-Intelligence: Living and Working with AI considers various facets of what it means to live in an AI-integrated world, including the ethical, social, and personal impact. Throughout the book, he interweaves the role of the human and emphasizes what he calls “co-intelligence.”
The idea of “co-intelligence” with AI isn’t new with this book. Sal Kahn, the founder of Kahn Academy, in his 2023 TED talk where he discusses how he’s using AI to develop personalized tutoring for students, talks about using AI to enhance HI (human intelligence), a concept I refer to often.
In his book, however, Mollick clearly breaks down this concept into four guiding principles:
Principle 1 - Always Invite AI to the Table
In my work, I like to give enough examples so that people can get a sense of what AI can do and some of the ways in which it could help them, but not enough that it feels like a list of choices or suggestions. Going with the idea of co-intelligence, AI usage can be very personal to the individual as it depends on the specific work and strengths of each individual. Mollick comments about innovation: “it is expensive for organizations and companies but cheap for individuals doing their jobs.” The approach that I suggest is to familiarize yourself with generative AI and its capabilities and potential and then to experiment. Think about some of the tasks you do with which you struggle the most and/or take you the longest and/or you dread and consider how can AI help with those.
Principle 2: Be the Human in the Loop
The concerns with using AI are real – bias, hallucinations, misinformation. Stuff is sometimes just wrong. Mollick explains that AI’s operating procedure is “‘make you happy’ beats ‘be accurate’.” AI so badly wants to give you what you want (because that’s how it’s been trained) that it will give you wrong information, if it thinks that will answer the question. In a world of co-intelligence, there is still a role (a big one) for the human. In my work, I tell people to consider AI like your 23 year old intern. You would never pass off work from the intern as your own without at least proofreading it. Same with AI. Always review the outputs you get, and (just like with an intern) give the tool feedback. It wants you to be happy, so it will take the feedback seriously.
Principle 3: Treat AI like a Person
This is perhaps my favorite principle because I chuckle every time I think about it. Mollick says, “AI is easiest if you think of it like an alien person,” and I think that’s such an accurate description of it because AI interactions are always just a little bit awkward – from the potential hallucinations and misinformation to the role of the training data (and limitations thereof), AI is definitely alien-like. Mollick adds, “(but tell it what kind of person it is).” In the co-intelligence relationship, you are the human, and you need to be the guide for the alien person. When I do trainings, I make sure to emphasize that the human needs to take control in the interaction. Tell the AI who it is (for example, an expert teacher), what you want from it, what the output should look like (for example, a blog or a chart or bullet points), and any other guidance.
Principle 4: Assume this is the Worst AI You Will Ever Ue
When I do workshops and professional learning with schools and organizations, I emphasize this point, saying, “AI is only going to become more powerful and pervasive.” We’ve seen what it’s capable of today in what is basically the beta versions of the technology. We’ve seen the improvement of ChatGPT 3 to 3.5 to 4 in a little over a year. We’ve seen how quickly the technology has evolved to be more accurate and more human-like. It’s not going anywhere, and it’s not going to get any less better. Approaching AI with co-intelligence in mind at this stage, allows us to grow together and develop together.
Yes, through a co-intelligence approach, AI has great potential to help you become more effective and efficient.
Yes, it can do really cool things that you can use to your advantage.
And, to answer your real question, no it won’t replace you. There will always be a role for a (co-intelligent) human.
Comentários