I love riddles. One of the first ones I heard as a kid was about the Liar and the Truth Teller. There are two guards and two doors. One door leads to freedom, and the other to death. One guard always lies, the other always tells the truth. They know which they are. They know where the two doors go. You don’t. You may only ask one question to figure it out.
What do you ask to determine which door leads to freedom? (Answer at the end of the post).
The average person makes 35,000 decisions per day. That’s not a typo. From the moment we wake up, we’re faced with endless decisions – big and small – that require our immediate attention.
But how do we make those decisions?
In the example above, you could have just picked a door, but you chose to ask a question to gain more information. When you have a decision to make in real life you take a bunch of inputs (information) and deliver an output (decision). If the decision is easy, inexpensive or repetitive, you probably use your gut. When you’re less confident and a decision is expensive (think risky or costly), you might seek more inputs in the form of information from news sites, blogs, a professional or friend.
Information is coming at us all the time on social media, on podcasts, from friends, from the news. In most cases, if you look for information or ask someone for advice they will probably give it to you.
But should you take it?
I have two friends who seem quite similar on the surface – work in the same industry, similar levels of education, etc.. Adam is highly intelligent and well read. He focuses on a few narrow fields and if he gets interested in a topic, he’ll read the footnotes to the articles and listen to a podcast about the footnote. After he’s analyzed a topic, his predictions will be right 98% of the time; when he speaks on something else, he will always qualify his opinion to let you know he’s not sure. In those cases he’s often wrong. Our mutual friend Sam is also interested in a lot of different topics and is pretty up to speed on current events. Sam will throw out an opinion on just about any topic without qualification. He speaks with the same 95% level of confidence about real estate finance, 16th century Russian art and pickleball. His accuracy runs the gamut from 90% on certain topics to 5%.
The big problem that is happening to you all day every day, is figuring out which inputs are worth including in your decision. Can you distinguish Adam from Sam? Do you have a partner, colleague or friend who sounds confident 100% of the time? How often are they right? How confident are you in their confidence?
Test the Inputs
Ideally, you would test these inputs with the actual results of the arguments and opinions. I’ve been trying to do this to myself for over a decade. I make a prediction (interest rates will go up this year, people will pay at a coffee shop using Google Glass by February, etc.) and I put a reminder in my calendar for some date in the distant future to test myself. This has helped me get better at prediction and at trusting resources.
As you go through life, you should continuously be calibrating your gut or ability to predict. After you see a few different scenarios that look like the one in front of you, these predictions will hopefully become more accurate. This is also happening when you decide whether to cross the street in traffic, buy a stock, take a new job, etc.
What is happening in the world of machine learning is similar. There are a lot of models spitting out a lot of data but often times people dont stop to understand the inputs. If you feed a model bad or biased information, it will give you a terrible output. ChatGPT is amazing, but do most people realize it is often inaccurate? What are you using it for?
Looking Forward
Now what does this all mean about how decisions will be made by humans and machines moving forward? As we head towards an increasingly automated future (and a world in which inputs and outputs become more readily available), it’s important that we find a balance between the capabilities of AI and the critical thinking of humans. AI is only as good as the data and algorithms it’s trained on. Humans, on the other hand, have the ability to think critically and make judgments based on personal experiences and emotions. Therefore, the key to successful decision making will be finding a balance between the precision of AI and the intuition of humans.
Hopefully, this will produce more Adams and fewer Sams. At a minimum, you want Adam around when you’re making an important decision, like buying that house. Sam can help you pick a good restaurant for lunch.
This ability to filter (or figure out who is Adam) might become the key determinant of human success in the near future.
The answer to the riddle: you ask either one “If I were to ask the other guard which door leads to freedom, what would s/he say?” You then do the opposite.
