Original Source Here
AI: How we failed The Marshmallow Test
Like a child taking the marshmallow test, decision makers have two options when dealing with questions they do not know the answers to. Only one of them paves the way for a bright future.
What if AI is not about finding the best answers to the most difficult questions, but about avoiding dealing with the fact that some questions cannot be answered?
Asking scientists and decision makers to deal with a question they do not know the answer to is like asking children to take The Marshmallow Test.
It is as uncomfortable for decision makers to have to wait for an answer as it is for children to have to wait for a marshmallow.
The second the question is posed, they want to know the answer — just like the child wants to eat the marshmallow the moment she lays eyes on it.
But just like there is a reward for the child who delays her gratification — she not only gets two marshmallows instead of one, she also increases the likelihood of getting a bright future — there is a reward for decision makers who dwell on the questions.
In fact, when decision makers endure the discomfort of not knowing the answers, they realize that answers are actually the least interesting thing about questions.
Before and after the marshmallow
Ask any child — or adult — and they will confirm that they felt way better before eating a marshmallow than they did after.
Before eating a marshmallow we feel excited and eager to do what it takes to be able to enjoy the treat. We use our senses to smell the marshmallow, and we use our imagination to picture ourselves eating it.
If we haven’t seen that kind of marshmallow before, we ask ourselves whether we think it tastes like strawberry or vanilla. And if there are other people in the room, we immediately invite them to explore our questions with us.
In contrast, we are neither excited nor eager when we have eaten the marshmallow. Instead of sharpened senses, we have a nauseating feeling that we probably would have been better off eating a carrot.
And rather than being curious and keen to connect with other people, we disappointedly conclude that nothing about the whole experience lived up to our expectations.
So what do we do about tough questions?
Like a child taking the marshmallow test, decision makers have two options when dealing with questions they do not know the answers to. Either they:
- Shorten the distance between question and answer as much as possible (in 2021 that means having AI do as many calculations as possible as fast as possible), or they
- Acknowledge that by shortening the distance between questions and answers they also shorten the time we use our senses and imagination to explore and discuss our opportunities
When decision makers endure the discomfort of not knowing the answers, they realize that questions have other and more important functions than reaching the answers as quickly as possible. Crucial functions like:
- Feeling excited and eager to do what it takes
- Activating our senses, imagination and experience
- Exploring different understandings of our problems
- Inviting others to take part in finding the solutions
- Collaborating on making the solutions a success
- Aligning on what is and isn’t important
This is exactly what most decision makers say they want, but it is the exact opposite of what they do when they drain their companies of excitement, curiosity, conversations and collaboration by replacing human interaction with artificial intelligence.
The beauty of being a “question animal”
There is no doubt that AI is useful and valuable in many contexts, but AI is designed to give us as many answers as possible as quickly as possible, and humans are designed for something else.
If decision makers don’t realize and respect that, they end up creating companies and societies that keep us in a constant state of nausea and disappointment where nothing lives up to our expectations.
The beauty of being what C.E.M. Struyker Boudier called a “question animal” is that not only are we able to adapt to change, we thrive on the openness and attention to opportunities that precede and define change.
About the author:
Pia Lauritzen, PhD is a philosopher. She has written several books and many articles about questions. She is also the inventor of a question-driven tech tool and the co-founder of Qvest — a Copenhagen-based startup that supports the transformations of purpose driven legacy organizations.
Trending AI/ML Article Identified & Digested via Granola by Ramsey Elbasheer; a Machine-Driven RSS Bot