What I am looking for as a young engineer

What should one seek out as a young engineer? In a phrase: real world feedback. By real world feedback, I mean the following: if I do a good job, then the thing works. If I don’t, then it doesn’t.

What kinds of work entail real world feedback? Any kind of work where you design or build a thing, and then watch it get used or tested. Types of work where no matter how you frame things, success and failure are both clearly defined.

This is important. If success and failure are not clearly defined, your iteration function is political instead of technical. Your success is contingent on the extent to which you can convince the people around you that you are doing a good job, rather than on how good of a job you are objectively doing. There is a forcing function for you to get good at talking, rather than actually being good at your job.

For high-integrity people, this is less of concern—having high integrity means that you will make the best effort at doing the right thing, even if it goes against your immediate incentives. High-integrity people don’t kiss ass, at least not consciously. But even if you are a well-intentioned person, if you do not get feedback from reality, you can still be misdirected! You could be spending your time doing what you think is the right thing, but is actually harmful—merely because the people around you approve of it. With hardware this cannot happen because even if everyone around you thinks the thing you are working on is a good idea, your system still must pass a test. There is a forcing function towards truth that exists outside of your social hierarchy. As Elon has said many times, “Physics is the law, everything else is a recommendation.” If you get no feedback from reality (physics), then you are only operating off of the recommendations of everyone around you. You aren’t learning how to build—you’re learning how to please.

Working for quality, I currently find myself in a paradigm where, in most things that I do, there is no clear or consistent definition of failure, and thereby of success. Nevertheless, since I am being paid by someone else, I must continue to demonstrate to them that their investment in me is worth it. In other words, my reward function is mostly political. I am navigating this by defining clear success metrics for myself, bounding them by time, and broadcasting them to others. This enables me to define a personal criteria for success or failure, a hard yes/no against which I can evaluate my own performance. In this way, I will build up a track record of executing against my own objectives, which I will then use as leverage to negotiate for more tangible responsibilities.

My ultimate goal as an engineer is to learn how to create things that are useful. This requires a practical understanding of physics that extends beyond the textbook and into real life. Reality is different from people: it works in a determinate way—no amount of convincing or charisma will change the laws of physics. Thus, with physics, we must take a different approach: we must understand the constraints they pose, and build within them. The best way to understand such constraints is to try to build something, potentially fail, and then try again until we find something that actually fits. Practical skills will enable us to solve many of the most important problems that face us in the 21st century: climate change, space exploration, water scarcity, etc. These are real world problems, and to solve them, we will need engineers who understand the real world.¹

1. There is almost always a political component to engineering. To build anything, you have to martial resources, raise funds, and organize people. While these skills are important, they are secondary to the practical knowledge that building something requires. To martial resources to build something you don’t have the faintest idea how to build is quite stupid.