Future AI Products Might Be Habit-Forming, In a Good Way - with Nir Eyal

Speaker 1

And contingency is this concept that I came up with where if you think about the contingent relationship between a mother and a child, that when a mother breastfeeds her baby, that baby as reaction with the mother lets milk down and then there's a contingent relationship where the mother produces more milk as the baby needs more milk, we will see this type of relationship with our technology as well. And this happens mostly in the investment phase of my hook model. We can talk about the hook model in a minute. But essentially it's this four part model where the fourth step is what's called the investment phase. And when I first wrote the book, this was back in 2014 and there was very elementary ways to get users to invest in the product data, content, reputation, skill acquisition. That's the part of the hook model that we are going to see blow up. Because what's going to happen is that the amount of contingency that companies will be able to deliver in real time, meaning consumers will change the product and the product will change the consumer just like that mother and child relationship will happen much faster. So products will be customized as we see generative AI and generative technology. You're going to see these products customized to your taste and preferences for a market size of one, right? Every product will be customized just for that user. This is really strong. A couple quick ideas on this and then we are going to move into the hooking and kind of habit forming elements of these products.

Speaker 1

They want you to use forever, but the only way they can do that is to create a healthy value exchange. Yeah, I think that is a force to consider the fact that Apple bake in these blue light things or these use limiting things. I don't know if it stops what some might argue would be a borderline inevitable trajectory into really immersive VR all the time, but it may. I mean, it very well may. I couldn't say it doesn't. Final thing we'll wrap on if we've got eight minutes or so for it would be some of the other forces on the technology side and your second book talks a good deal on the behavior side where it's somewhat crunch on time. When you think about the tech components, maybe the uses of AI and other virtual technologies to bend us towards doing what we want to do. Like you said, maybe watching football is what you want to do and that's okay. Maybe really cranking your sales numbers what you want to do. That's okay. What are the technological forces pulling us in terms of our own autonomy, promoting our what the heart appoints for us and keeping us in line with that? Yeah. So I can see a future where there is an agent, maybe some kind of conversational interface that asks you ahead of time what your intentions are. So one of the techniques that I describe in a distractible which has been studied thousands of times in peer review journals is called setting an implementation intention.

Speaker 1

So imagine if there was a technology, this has always been the missing link. Imagine if there was a technology that preemptively could see that you're becoming bored, that you're becoming anxious, that you're becoming stressed and knew that you were likely to get distracted, right? This is the hardest part about teaching people to become indestractable is noticing those preceding sensations that lead us to distraction, that lead us to Facebook, that lead us to the cigarette, that lead us to cheat on our diet. If you can bring awareness to that. So for example, let's give a very concrete example. I am certain that one day there will be a product that's, you know, maybe the size of this button on my shirt that looks at what I'm about to eat. And before I eat it, I calculate the nutritional content of that food. And so as opposed to me logging my food after I eat it, which is very difficult to get people to do, very, very high friction, instead this little device, you know, tells me, hey, that plate of food has 1200 calories. If you remove those French fries, it's only 400 calories. And before you eat it, we just wanted you to know, like if it could preemptively tell you information that changed your behavior, because right now most behavior change applications, they're all retrospective, right, they're lagging indicators. The future will be a future filled with leading indicators. If you go to sleep right now, this will happen. If you eat this versus this, that's what's going to happen versus today.

Speaker 1

If you can bring awareness to that. So for example, let's give a very concrete example. I am certain that one day there will be a product that's, you know, maybe the size of this button on my shirt that looks at what I'm about to eat. And before I eat it, I calculate the nutritional content of that food. And so as opposed to me logging my food after I eat it, which is very difficult to get people to do, very, very high friction, instead this little device, you know, tells me, hey, that plate of food has 1200 calories. If you remove those French fries, it's only 400 calories. And before you eat it, we just wanted you to know, like if it could preemptively tell you information that changed your behavior, because right now most behavior change applications, they're all retrospective, right, they're lagging indicators. The future will be a future filled with leading indicators. If you go to sleep right now, this will happen. If you eat this versus this, that's what's going to happen versus today. It's all lagging indicators that already happened. It's too late. You know, it's nine o'clock and at night, your, your little Fitbit says you've only walked 6000 steps, but you need 10,000. Well, I can't do anything about it. It's too late. As opposed to in the future, it'll have a leading indicator to say, Hey, do this now because this is going to be the result later.