There is something you could call engineering alpha or stack alpha. For eg knowing a language exploding in demand. But that’s a degenerate case. The full effect relates to stack maturity. The more tools you have to know to build a complete thing, the lower the stack alpha.
Conversation
An immature stack like Web3 might have only a few layers. That’s high alpha, and you can build rough frontier things with very high upside potential. What you build will mature along with the stack itself, enjoying multiple rising tide effects.
3
1
13
A mature stack like Web2 might have 3x to 4x as many layers/tools/moving parts to build a complete thing. And while you can command huge salaries and stock at mature companies if you’re in the top 1% of stack mastery, the median stack-master will make much less.
Replying to
This gets at something that’s been bothering me for a while. The depth/breadth of technical knowledge needed to do a job do *not* correlate well to compensation. For example housing general contractors on HGTV shows seem to be walking oracles, but it’s a tight margins business
3
1
14
There’s a product dimension too. For example if a new language becomes popular for say robotics, and you put in serious hard work writing or porting a linear algebra library for it, everybody will love you and thank you and you’ll be viewed as a mensch, but not get much reward.
1
18
But if you do something that hooks into the alpha of the new stack in a fundamental way, for eg in the case of a robotics language maybe the best concurrency module… you’ll get all the esteem AND financial rewards. You’ll probably end up as a senior architect at a top company.
2
1
13
The Q-James Bond relationship is an interesting example. Q invents new stacks, but Bond usually reaps the rewards because he can cash out the alpha effectively. More true of old hardware-Q than young new hacker-Q.
2
1
9
Yep. Unfair, but true. Tech is a business that rewards cashing out stack alpha with frontier product engineering way more than keeping critical things running in civilization core.
Quote Tweet
1
6
26
Engineers I meet naturally seem to fall into two types:
Tool-master types who excel at all pieces of a mature low-alpha stack
Strategic technologist types who are often mediocre at the tools they use, but have a nose for seeking out and leveraging high-alpha stacks
1
5
23
The latter usually log big wins early in career and establish reputations as “technical entrepreneurs” so to speak (ie they wrangle stack risk, not market risk) and end up in engineering leadership.
The former usually end up as senior fellows or something if they’re good.
2
11
Engineers don’t seem to talk much about risk management from an upside perspective. Only downside. Eg security holes in a fragile new stack. But some seem to have a natural instinct for upside. As in “if I use this janky new component now, it will pay off big time in 3y.”
3
2
23
By the time something becomes a “tech trend” that Gartner is writing about, much of the alpha is already gone. But sometimes not. For eg. machine learning has been a flagged trend for years, but because of hardware being the bottleneck, much of the alpha is still up for grabs.
1
8

