First principles and technology
Several years ago, I wrote that we're collectively building the equivalent of an Iron Man suit, but each of us within the suit is a lesser form of those people who existed in the past.
Lately, I've been thinking about how that applies to technology and specifically computers, which most would agree are the most important devices to mankind today. There's a problem though – nearly everyone we know can use a computer (or at least its modern form, the smartphone), but very few of us actually understand computers. While there are probably more computer scientists today than there has ever been, the ratio of computer scientists to people who interact with computers every day has never been lower. In other words, almost everyone today uses computers, but an extremely small fraction of these users actually understand them.
Step into any office, and you would be hard pressed to find someone who can troubleshoot a computer's networking issues, never mind specific software or hardware issues. Outside of perhaps PC gaming enthusiasts, few people own an actual computer these days – and those few who do are likely to just own a MacBook laptop that they can't even upgrade (I'm guilty of this).
Take this even further, and ask – how many people actually understand the first principles of computers, such that they could reconstruct a computer from nothing if they had the resources and the time to do so? It's truly a scary thought to know that if some calamity were to befall mankind, huge swathes of technological advances could be lost because knowledge of the associated first principles is concentrated in just a few individuals.
One could argue that first principles may be becoming less important in certain areas. For example, Jensen Huang recently stated that before long, computing technology will imbue everyone with the ability to program. On the surface, it seems that he means we will be able to program in plain English. This made me think – even if we could build applications using regular sentences, would that make programming any more accessible to the typical person?
My (admittedly limited) experience with ChatGPT has shown me the importance of prompting correctly. If we don't have any background whatsoever in algorithm design or databases, or any understanding of distributed systems and networking, could we really create anything of value that supersedes what others have done before us? On the flip side, it seems to me that people who understand computers deeply will become even more valuable in the future, as they could harness AI tools to replace junior programmers in the tech ecosystem.
Given the above, my hypothesis is that first principles are as important and relevant as ever. To borrow from Naval Ravikant's wisdom, topics such as discrete math, microeconomics, logic, and physics will continue to hold value. Reading books will not go out of style anytime soon. Writing this post is a reminder to myself of these conclusions.
