How AI Is Reshaping The Prisoner’s Dilemma
By Gareth Henry | Last Updated 21 December 2022
AI might be the key to solving the Prisoner’s Dilemma, but we won’t know until we’ve tested it out on our least favourite family member (That joke was generated from AI!)
In all seriousness though, your immediate society peeps won’t know what hit them when your AI driven game theory strategies come to the fore.
But before we get too excited…
What Is The Prisoner’s Dilemma And Why Is It A Problem?
The Prisoner’s Dilemma is a classic game theory problem in which two people are given the option of either cooperating or defecting in a situation where their individual interests are in conflict.
Here goes the often-used Prisoner’s Dilemma scenario:
Let’s take two criminals, Adam and Ben, who are involved in a robbery. They’ve just been arrested and placed in jail. They can’t talk to each other, so they need to make all their decisions on their own.
The prosecuting lawyer offers each prisoner a deal. There are four possible outcomes in this scenario:
| Scenario | Outcome |
|---|---|
| Adam & Ben both snitch on each other | Adam & Ben each serve 3 years in prison (6 years collective) |
| Adam snitches on Ben but Ben doesn’t snitch | Adam is free, and Ben serves 5 years in prison (5 years collective) |
| Ben snitches on Adam but Adam doesn’t snitch | Ben is free, and Adam serves 5 years in prison (5 years collective) |
| Adam & Ben don’t snitch on each other | Adam & Ben each serve 1 year in prison (2 years collective) |
Remember, Adam and Ben can’t talk to each other, so they have no idea if the other is going to snitch or remain silent.
Individually, Adam and Ben may believe that the best outcome is to snitch on their friend and hope their friend hasn’t done the same. They might be able to walk out of jail completely free.
But! By acting in their own self-interests and snitching on their mate, they will surely both suffer the most with a combined 6-year sentence when both decide to cheat.
The best strategy collectively is if they both remain silent.
As you can imagine, in practice this is a hard strategy to pull off when someone is dangling freedom in front of your face.
But, if we can’t believe we can achieve a better outcome for ourselves and our family and friends through collective action, that thought quickly becomes reality.
Striving to optimise our own payoff regardless of the larger group’s interests will ultimately result in a suboptimal solution for everyone, in nearly all walks of life.
How Can AI Help Us Explore The Prisoner’s Dilemma?
Now that AI is involved, we can explore the Prisoner’s Dilemma in new ways.
Firstly, AI can be used to analyse data from given scenarios. By doing this, AI can help to identify patterns in the data and provide better predictions for the outcomes of each decision.
Let’s take the example of Adam and Ben again.
When the lawyer is offering up the deals to Adam, you can bet that he’s only hearing how the offers sound from his point of view. He’s more likely going to act in his self-interest because he’s not looking at the bigger picture.
What if Adam and Ben were able to use AI? The AI would be able to analyse the Prisoner’s Dilemma in less than 60 seconds and reveal that statistically the best scenario for both is to remain silent.
Secondly, AI can provide more detailed feedback on the decisions made, allowing for a better understanding of the implications of the decisions made.
Sure, it’s great to make AI driven decisions. But knowing why the AI came to those conclusions is where the gold lies. This could potentially lead us closer to finding strategies that could guide real life solutions for some of our most frustrating problems.
How Can We Use AI To Solve Our Daily Prisoner’s Dilemma?
Let’s be honest. In our day to day lives, human interactions can be tiring and lead to decision fatigue. AI can help us muddle through those decisions so that we can have a clear mind.
We no longer need to make emotional calls and turn ourselves into a pretzel throughout the day as we make decision after decision.
AI can help identify and solve our daily dilemmas by providing insights into how we can best reach a cooperative outcome by evaluating everyone’s incentives. These recommendations could give people the confidence to be more willing to collaborate, knowing that their decision is backed up by science-based analysis.
Not sure if you should attend after work drinks? Can’t decide whether to include a slide in your big presentation PowerPoint? Making AI driven choices can help take those tasks off your plate so you can focus on the things that really matter to you.
Ultimately, appropriate use of AI may make difficult decision-making easier and simpler while promoting lasting harmony and support between all parties involved.
Where Will AI Adoption Take Us In The Future?
With the rise of AI, the concept of the Prisoner’s Dilemma has been revolutionised. What was once a mere thought experiment can now be applied to real-world scenarios, allowing us to explore the potential of AI adoption and its implications in a variety of contexts.
AI can be used to help us better understand the consequences of our actions. By utilising AI, we can gain insights into how our actions will affect us in the future and how we can best prepare for them. This can help us make better decisions in the present and ensure that our decisions are in line with our values and goals.