You Are Optimizing for the Wrong Metric | Lucas Pauker

Table of Contents

We love to optimize

Money (is the root of all evil)

Should you get good grades?

So what?

We love to optimize#

Technical people love to optimize every little thing.<br>For example, I am always thinking about how to use my time, money, and knowledge in the most effective way to the point of getting upset when I do things suboptimally.<br>However, everyone optimizes for different things.<br>When I had my first software engineering job, I was learning to use vim and I downloaded a bunch of different plugins to make the app prettier and have more functionality.<br>Then, when I finally had my setup perfect with 10 different plugins downloaded, I saw that my boss was using stock vim!<br>How can you get anything done without neotree??<br>He viewed vim as simply a means to an end. How much do plugins really help productivity? Was I really optimizing for the right thing by using my time to learn these plugins?

Optimization in general involves finding a maximum or a minimum of some metric.<br>Usually we are either trying to maximize utility or profit or trying to minimize costs or a loss metric.<br>When we use optimization to make decisions, the metric becomes the most important choice.<br>By choosing the right metric, we can grow our businesses or make better decisions in our personal lives.<br>But choosing the wrong metric often leads to unintended bad outcomes.

Money (is the root of all evil)#

Money is a metric that companies and people often try to optimize for.<br>It is easy to optimize for money since you can count it and you can earn it in many different ways.<br>However, whenever money is the only metric that one tries to optimize for, it leads to disastrous unintended outcomes.

Social media companies make more money the more time that people spend on their apps.<br>Therefore, the way to make the most money is to make the most addictive app possible and show you the most ads.<br>This leads to Facebook, Instagram, and other social media pushing content, ads, and AI slop into the feeds of people who use their apps.<br>About one fifth of your feed is ads…and it is getting even worse with ads displayed in influencer content.<br>So is this good for society? Is this what people want?<br>No. This a result of optimizing for the wrong metric.<br>There are studies showing negative consequences of social media addiction.<br>For Facebook, having users spend as much time on the app leads to more money, so they don&rsquo;t care about the ethics around how they incentivize users to spend more time on the app.

This is true in other types of companies as well.<br>Even in trading, where the main objective is to make money, when money is the only metric of success it leads to bad outcomes.<br>For the past few years, Jane Street allegedly manipulated the options market in India to make hundreds of millions of dollars, money that was mostly lost by retail traders.<br>Firms like Jane Street play a vital role in providing liquidity in markets around the world, but there is a limit to what should be done to make another billion dollars.

Another example that shows the dangers of optimizing for money is healthcare.<br>If doctors were purely profit motivated, they would have no incentive to cure any patient.<br>In fact, they would have an incentive to provide subpar care keep the patients coming back.<br>Therefore, in order to be a moral and good doctor you need to have motivation outside of just making money.

From these examples, we can see that optimizing for money often leads to hurting the end user after a certain point.<br>Therefore, money is always the wrong metric to optimize for.<br>It can be an important metric, but it can never be the only one.

Should you get good grades?#

Another place that typically optimizes for the wrong metric is schools.<br>Students are pushed to optimize for good grades.<br>By pushing students to get good grades, the incentive is to get the right answers and do good on tests as opposed to actually learning.<br>I have personally tutored many kids and young adults in math and computer science.<br>Often, when kids are struggling with learning a math concept, they will just try to answer the questions correctly.<br>Their objective shifts from learning a difficult concept to just trying to guess the answer.<br>However, as a tutor/teacher I don&rsquo;t care if they get the answer right, I just want them to learn the concept!<br>This also happens when trying to learn new concepts without having a full grasp on the prerequisite ones.<br>Instead of understanding the new concept, kids are often forced to memorize algorithms to get the right answer.<br>This leads to worse outcomes because now they don&rsquo;t understand two concepts instead of one.

This problem will only get worse and worse with AI.<br>With AI, the right answer is only a prompt away.<br>If schools continue to only use correct answers as the metric for success, students will game the system and not learn anything.<br>Some good alternative solutions may be doing oral tests...