Leveraging comparative advantage in life

I had a shower thought recently, triggered by thinking about how I’ve grown in a professional context over the last 11 years.

Something that I was told (and believed) for a long time when I started working was to ‘work on your weaknesses’. In my case, the professional weaknesses were pretty obvious and (at the time) significant, so this made a lot of sense to 23 year old me. I have always been quiet, introverted, awkward - whatever you want to call it. And working in the media industry just makes those characteristics stand out even more, since 90% of the people you work with are at the other extreme of the scale. This is more of an issue in social situations to be fair, but from a work point of view, and specifically the work I do, it’s also very important to be able to ‘sell’ what you do. For me that means presenting, storytelling, and doing these things with some level of ‘charisma’, which doesn’t tango well with the quiet, introverted and awkward person from the top of the paragraph.

This is why my weaknesses were very easy to see at work - I was regularly put in positions where I had to present and sell my work to clients, and being the person I am, I wasn’t very good at it. And so the feedback that I got from colleagues, fairly, was ‘work on your weaknesses’, because they were so tangible and directly impacted my work that it seemed obvious. And I did work on them! There was some professional help in there through workshops and trainings, but it was mainly through repetition. If you do something enough times and are genuinely motivated to improve, you’re bound to get better at it just through direct feedback. After hundreds of attempts, today I’d like to say I’m a competent presenter - I won’t win any awards, but I’m definitely good enough to get my point across and I’ve had enough sales meetings end successfully to know that I’ve reached an acceptable threshold of capability in this area.

Now the question is - even after improving my presentation skills, I’d say they’re still a weakness relative to my other skills, so should I keep going? And that is the thought that triggered this essay: all of us have a limited amount of time and effort that we can expend over a given time period (let’s say a year). Today, I could spend that time and effort trying to get even better at presenting, but is that worth it? Am I better off spending that time somewhere else?

So what should I do?

The logical answer is go completely the other way and focus only on your strengths. Spend all your time and effort on getting better at whatever it is you’re already good at. We all have a sense of what this is for us in the professional world. For most of us, it’ll be related to our job function, especially those with some years of experience in the bank (time ends up telling you what you’re good at and you usually end up moving into a job function that supports it naturally).

This was my first thought - I’m good at data and programming, so just keep doing that. This ties nicely with a concept that should be familiar to the economists reading: comparative advantage, which basically says that in a free trade market, you are better off specializing in what you are good at. The important clarification is that it doesn’t mean ‘good’ relative to others, but good relative to your own skillset. Even if you are somehow worse than others at everything in the world (which isn’t true, we all have talents), it still makes sense in a free market to do what you are personally best at and then exchange your talents with others who are good at whatever you’re not good at (either directly by swapping work, or indirectly by paying them, befriending them - whatever works).

On the other hand, in doing some research for this post, I learned of an economist called Ronald Coase who pointed out in an essay that companies exist because outsourcing was expensive: finding people to do a job, negotiating with them, checking their work, ensuring secrecy, and so on. In this environment, even if you’re not really the best at something, it might be worth just doing it yourself to avoid the costs (financial and otherwise) associated with getting the best resource to do it. An example in the professional context could be writing emails: I might not be very good at it myself (this is hypothetical, I’m actually quite proud of my email-writing skills), but is it even remotely worth going through the required effort to float an RFP and outsource it? Probably not.

Today with AI this changes a little bit - a lot of the friction and costs that Coase talked about in engaging external parties are gone. If you need someone to do the best weekly data analysis, you don’t need to look for companies, send out an RFP, evaluate them, negotiate rates with them, and then start working with one. You just need to open Claude/GPT/Gemini and type what you want!

So if the friction of outsourcing is gone, the natural idea that follows is to really only focus on your strengths and outsource everything else to AI - after all, it’s already smarter than the average human in basically every field there is, and only going to get better. So embrace the comparative advantage idea - thanks to AI, it’s basically a free market and you can get someone else named Claude to do whatever you don’t want to do yourself.

Easy right? AI has solved everything - what’s the catch? Well, AI has definitely removed a lot of the costs associated with outsourcing that Coase talked about (RFP, evaluating vendors, negotiating, etc.) but it didn’t remove all of them. There are still two significant costs to using AI (on top of the obvious financial cost of the task itself which I’m ignoring):

Explaining what you want - when delegating to AI, this needs way more nuance and effort as compared to talking to a human. And if you still use AI with a keyboard instead of voice (something I want to write about soon), then good luck.
Actually evaluating what the AI does - is it what you wanted? Is it good? Are there gaps? What are the gaps, and what extra information does it need to fix them?

These costs are pretty real, and I’m sure many of us who have used AI to help with work can relate to them. I can’t realistically just outsource everything I don’t like to AI. I’m going to end up spending all my time explaining what it needs to do and then checking its work, because it’s not good enough yet to just let loose on its own (and for some of us it probably never will be).

There’s another reason to not outsource something that comes from the analogy of countries trading. Some capabilities are strategically important, and need to be nurtured locally no matter what the market indicates. In my case, presenting was the example that fell into this category - I literally couldn’t ask someone to come and present for me in my meetings, so I had to develop that capability myself.

So to recap - we have our strengths, which we decided to focus on, and some things that can be outsourced: anything easy to explain and easy to judge.

What about the rest? I think capabilities that fall into this category should still take up your effort, but only to a certain threshold. If you currently fall below that threshold, give them time and improve yourself, but don’t go crazy trying to become an expert if it’s not what you’re already good at or interested in. To stick to my presenting example - it was strategically important to my career, and I literally couldn’t hand it to anyone else, human or AI, so I worked at it up to a threshold that was good enough and now I’m good to stay where I am. Eventually, once you have a good enough understanding of the concept, this can be delegated to AI as well, because you will know how to describe what you want and evaluate what you get back.

This part is very important - it’s okay to lean on AI even for things that are strategically important, but only after establishing a baseline level of knowledge yourself. AI will have absolute advantage over you here - it knows more, can solve better, work for longer, but it needs direction and it needs steering, and that’s where you add value.

What can you do today as a professional?

Identify your:
- Strengths (programming, strategy building, creative concepts)
- Weaknesses that are important to your career but can be delegated eventually (legal review, PPT creation). Think about what the ‘minimum acceptable level’ of competence is in these areas for you to get to.
- Weaknesses that cannot be delegated at all (relationship building, presenting, storytelling). Think again about the minimum acceptable level of competence that you need to get to in these.
- Things that you’re not bad at but don’t enjoy, that can be delegated today.
Use your available time and effort to focus on your strengths and bring up the weaknesses to a minimum threshold. Let everything else be outsourced - today, it’s up to you whether this is to a human or to AI. All you need is to be able to direct and evaluate what you outsource at the required level of competency.

This way you can use your finite resource (time and effort) at maximum leverage and gain maximum economic value from your time.

So this is where I’ve landed. I know the next few lines sound like AI but I swear they aren’t.

Focus on what you’re relatively good at. Good means relative to yourself, not to others.
If you’re bad at something that’s easy to explain and easy to evaluate, outsource it.
If you’re bad at something that’s hard to explain, hard to evaluate, or genuinely needs a human, work on it until you reach a baseline. Then stop.

The constraint in life is your time and effort, so it makes sense to spend it where you should, not wherever you can. Find your comparative advantage and go all in. Yes, AI has absolute advantage over you in basically everything today, but the point isn’t to prove that we’re better than AI. It’s to leverage it to use your time on this planet in the best way possible.