The flaw in this article is the assumption that 10x engineers are just more productive, and several "normal engineers" can do the work of a 10x engineer. This may be true if you're building "normal software".
But for certain kinds of software development, a team of "normal engineers" can't do what a single 10x engineer can do. For example, how many "normal engineers" would you need to replace an Ilya Sutskever? The answer is you can't. Because intelligence is not stackable.
How many companies out there genuinely need an Ilya Sutskever to achieve their goals? Everything in this article is accurate for the 99.9% of companies and teams that aren't working at the bleeding edge of the industry. The mythical "10x engineer" is always a net negative on teams building a boring CRUD app. You always want a handful of "normal" engineers instead.
Intelligence is the easy part. That’s not what makes a 10x engineer. The 10x comes from consistency, persistence, creativity, resourcefulness, patience, focus, communication, drive, empathy, and a bunch of other traits that are rarely found all together in a single individual.
No, intelligence is not the easy part. We are talking about uncommon intelligence and it is the hard part. And they ones that have it are often lacking in other areas, such as social skills. And the ones that have social skills, and other traits often aren't 10x.
Put it this way: of the handful of 10x engineers I’ve met in my life, it was never intelligence that was the differentiator or what impressed me. That’s a commodity as far as I’m concerned. Once you pass 140 IQ it’s mostly diminishing returns, compared to the outsized impact of the other dimensions. It’s always been the persistence and consistency that has stood out the most to me.
> For example, how many "normal engineers" would you need to replace an Ilya Sutskever?
I wonder what Sutskever would think if you asked him that question. I bet he'd point out several of his contemporaries that are more deserving of praise.
I'd argue there are plenty of Sutskever's out there who are normal engineers.
There are plenty of hard problems that are unrelated to scientific research where you need more than just "normal" engineers. Look at the number of failed projects at big companies such as Amazon and Microsoft, they failed because "normal" engineers couldn't pull it off.
hah. i'm sorry but this is extremely naive. projects don't succeed of fail solely on the technical chops of the engineers. it's a whole "ecosystem" that has to work and most projects I've seen failing are due to politics/bikeshedding at upper management level.
Your experience is different from mine. I have worked at big tech companies and have seen too many projects fail due to lack of technical chops. Do they fail for other reasons too? Sure, such as not finding product-market fit, but at some companies they start with guaranteed product-market fit because they are copying a competitor!
> We place too much emphasis on individual agency and characteristics, and not enough on the systems that shape us and inform our behaviors.
This is just author trying to impose their personal politics on their org. Popular sentiment among pmc unfortunately especially in the last 10 years of zirp
Systems thinking isn't a political belief, it's a model for the world that is incredibly useful in a lot of contexts. Some political factions may be more likely to engage in systems thinking than others, but that doesn't make it a political topic.
my rule of thumb is that the average startup has no more than one novel piece of code, and it is usually one of the first things written. Then after that there’s enough plumbing and UI work to keep a few hundred regular engineers busy
But for certain kinds of software development, a team of "normal engineers" can't do what a single 10x engineer can do. For example, how many "normal engineers" would you need to replace an Ilya Sutskever? The answer is you can't. Because intelligence is not stackable.