How to make optimal decisions part 1: goals
Having reached a point in my life last year of any combination of diverging paths in many yellow woods, I thought it was worth investing in learning how to make decisions. After all, I’m expecting to meet quite a few of them in life. ;) This blog post is the start of a series summarising the most useful pieces of advice I’ve found from my MSc in Human Decision Science and a range of other sources. This one starts off a bit philosophical but the ones to follow offer very practical pieces for making optimal decisions in the small stuff like how to organise your schedule to bigger stuff like choosing a career. A previous blogpost of mine ‘Choosing a partner according to decision science’ gives a flavour of things to come but is a little more tongue-in-cheek.
- Choosing a goal
As a general starting point, it’s worthwhile thinking pretty seriously about your goal. A common problem is optimising the wrong goal. Take, for example, the career parent who optimises their career by using every hour of the day to work. In doing so, they neglect their kids and so don’t maximise their actual goal of providing the best life for their kids (I’m hoping that there is enough general agreement that parental input is important to child development that I don’t have to find a citation!).
I think this is even more important in the current times where we have access to lots of data and a culture increasingly worshipful of evidence-based decision-making. This creates conditions for a very powerful drive towards the goal, which is great… if the goal is what you actually wanted. In a now infamous thought experiment, AI researchers thought through the consequences of creating a machine whose goal was to optimise paper clip manufacture and was capable of improving its ability to do so. Without the appropriate checks and balances, the whole world could be covered in paperclips, a goal orthogonal to goals many would consider more important such as sustaining human life on the planet. The Fun Theory Sequence is a really interesting exploration of the big picture goal of achieving utopia. It suggests that actually we’d get bored pretty quickly with what most people think of as heaven.
2. Defining your alternatives
So how do we choose a way to achieve our goals then? One way is relaxing the constraints on the problem. This idea underlies advice to ‘think outside the box’ e.g. ‘what job would you do if all jobs paid the same?’ But then it’s crucial not to forget the step of translating it back into reality: living each day as if it’s our last is not actually conducive to long-term happiness: you’d never sow the seeds of great achievement (which I believe would be conducive at least to my personal happiness!) but it’s useful in highlighting the things that are important to us that we often forget.
Another idea which is particularly salient to me at the moment is the idea of local maximisation. I’ve just accepted a pretty cool job in my hometown, which is likely to be reasonably close to the local maximum, the best fitting job I could hope to find in my hometown. BUT that means I could be missing out on some global maximum, the best fitting job I could hope to find in the world. It’s easy to get stuck maximising locally and forget that there are so many other options. So start by listing as many options as possible even if you think they are reasonably unlikely to come off (for example, living in Copenhagen and learning Danish made my list of next step options). Then randomly picking another locality (in this case city), find the best option there (in this case a job in Copenhagen) and see if it beats the current global winner. (Here I’ve very crudely adapted a Shotgun Hill Climbing algorithm — computer scientists please forgive me; for those interested see ‘Algorithms to Live By’.)
3. Comparing options
One way is to create a list of criteria and score the alternatives against these criteria. This is how ‘Which?’ magazine evaluates products. However, it’s worth keeping in mind that we’re not very good at knowing what we want in the long-term or makes us happy i.e. at choosing the right criteria. This is why I particularly like the approach of 80,000hours which has developed a decision tool to help early career individuals with career decisions: they suggest criteria to evaluate a choice based on the happiness literature from psychology. For example, helping others is robustly associated with life satisfaction and so they suggest it as a criterion to evaluate a career. (You can tell careers are somewhat on my mind at the moment! The job is the biggest decision I’ve had to make of late…) For evidence-based happiness advice I particularly like Action for Happpiness which has members in 170 countries: having such an appeal across different cultures and faiths suggest they’ve captured something somewhat central to the human experience.
However, keep mind that that the happiness literature gives you what makes people happy on average: there’s still quite a lot of variation which isn’t captured. For example, the part of jobs I’ve always enjoyed the least is the admin but I’ve had previous colleagues who love it for the sense of accomplishment when they complete it. Psychological studies suggest that we remember the beginning, end and most intense parts of an experience, and so turning to past experiences to help decide what criteria to evaluate can be misleading. We also underestimate how much we as individuals change — my least favourite subjects at school were maths and IT and here I am using statistics every day and learning to code. We also habituate pretty quickly: that apartment that is going to transform your life? You’ll be worrying about other things soon enough.
One way to overcome these biases is to track your own happiness yourself: there are apps nowadays that help you do so. One I particularly like is Hippo which prompts you to record your happiness (including different dimensions of happiness such as joy and a sense of purpose) at random points in the day, and what activity you’re doing at the time. The slight downside of such an approach is that our experiences are not independent of each other. If I’m worried about an argument I had with my sister, that will colour otherwise enjoyable activities. But hopefully these confounders even out over time… Then you have the fun of experimenting a bit: increasing the activities that bring you happiness, and decreasing the activities that bring you down.
In the last paragraph I’ve assumed that happiness is what most people are aiming for. I think it’s generally fair to assert that people seek positive feelings and experiences but I don’t necessarily mean a state of constant ecstatic joy. In a recent article I really liked, the author claimed she was content with being unhappy but interested — being interested in things gave here a sense of purpose and helped her make meaning out of her experiences. Happiness is pretty multi-dimensional and so if you can be even more specific about the type of happiness, then that’ll help you better identify the criteria on which to evaluate a choice.
It’s quite exciting to think that there are optimal ways to make decisions about lots of things, and I started the MSc in Human Decision Science partly expecting to be given answers about how best to live (I have high expectations I know!). But as I’ve tried to get across in this post, the goal is the most important aspect and that is entirely up to the individual. So what would you like to aim for? :)
