Engineering Teams Are Just Networks | Hacker News Reader

Engineering Teams Are Just Networks

bellmar.medium.com

65 pointsmbellotti24 comments

These articles always skip over the all important context: What is the engineering team supposed to accomplish?

> After a baseline level of competency is satisfied, who you hire does not matter.

For your average CRUD web app this is obviously true. For the Manhattan Project it’s obviously false. Most engineering work falls somewhere in-between.

Funny you bring up the Manhattan project, as the actual uranium was purified by the "Calutron Girls".

https://en.wikipedia.org/wiki/Calutron_Girls

These were high school graduates (largely female), who sat around fiddling with the controls to manually run the uranium purification knobs.

Apparently, engineers and scientists were really bad at this job. I don't know how they figured it out, but high school females had just the right mindset to watch the control panel and move the dials correctly.

None of the Calutron Girls knew they were making the bomb. In fact, the article where I read about them was an interview with an old-lady in the 2000s, who was surprised to learn that her high-school summer job was building the A-Bomb!! I think they were told that it was very important for the war effort that they do their job correctly, but that was about it.

------------

So even for an incredibly complicated job like the Manhattan project (which obviously needed top-tier scientists), it turns out that top-tier scientists / engineers simply don't have the mindset to do some dreary day-to-day tasks. Other people (such as the Calutron Girls) may end up having a better mindset to do these parts of the job.

I wouldn't be surprised if our understanding of human psychology, sociology, and group-dynamics are still poor today. The ideal team composition is still a big mystery, especially on big hypothetical jobs like a "Manhattan Project".

Yep, before we had mechanical and electronic computers, there were human computers, and they apparently too were mostly women: [0] . I think (but it's just a speculation of mine) that it's not because women inherently have a "better mindset to do some dreary day-to-day tasks" but that it's an acquired trait: I blame it on heavy training in cleaning and sewing.

[0] https://en.wikipedia.org/wiki/Computer_(occupation)

That very well could be it.

I probably should be more specific: the enrichment facility had roughly 20,000 employees. That's more than enough time for managers to figure out who was good and/or bad at the job, as well as to build stereotypes / culture for how to pick out a good candidate for hiring.

It very well could be that these women had tedious hobbies (such as sewing, crochets, or other textile work), which made them predisposed to passing whatever tests the enrichment facility had.

In either case, the end result is the room full of tens of thousands of young women.

---------

Another link on the Calutron Girls: https://www.energy.gov/articles/five-fast-facts-about-calutr...

As far as I understand the author, it's always obviously true. She just ups the baseline to genius level if you're on the Manhattan project.

Which makes the statement mostly tautological: Hire only people capable of doing the work.

Maybe she doesn't want you to hire overqualified people, and wants you to shift focus on what qualities the team as a whole is missing.

All of this never makes it out of hiring 101 teritory, of course.

Update: s/he/she/

I mean, is it true even then?

I know it doesn't work like that, but as a first approximation we can imagine that people working on the Manhattan project would be able to write some javascript or whatever given the chance, while the opposite isn't true. So you would still want to hire the best people you can get your hands on: maybe you underestimated the difficulty of the current project, or the next one will be harder. Anything else is just Pareto-inefficient.

"You can get your hands on" is the imperative part.

All the article is trying to say (as far as I can tell) is "do not wait for the perfect hire if you have a perfectly good hire".

The "complex contagion" and "network effects" are what triggered the insight for them, but are not really necessary to understanding the principle.

So, it's obvious that if there were some people much smarter than the ones working on the Manhattan project, they wouldn't be able to improve it on any reasonable way?

I could maybe agree that it's true, but I really don't agree that it is obvious, and don't agree that it generalizes to all less demanding projects done by less capable people. It probably is true for a lot of things, but claiming it applies to all seems ridiculous.

“She”, I think.

Is that true of the Manhattan Project? The fundamental science that the bomb was potentially possible had already been worked out (which was why they launched it in the first place). Most of the work there was physically building the bomb, acquiring and working with the raw materials, performing tests, etc.; most of the complex work was still working from what scientists had already written.

Consider that it took Newton and Leibniz to invent calculus, but, once invented, we expect high school/college students to be able to learn it.

I think for almost all roles in the Manhattan Project (possibly everyone except Oppenheimer himself, and there were plenty of other candidates for even that role) you could hire and train anyone with a baseline level of competency in the work needed. It wasn't the sort of work that required genius, but at the same time, it was novel; nobody came into the project deeply experienced in refining uranium or whatever. So as long as you structured the project in a way that people could get up to speed quickly you were fine.

Given how few countries have managed to launch a successful nuclear program, in spite of the basic theory of how fission and chain reactions work being published a century ago, there must be something beyond basic follow instructions ability required. Though whether that is anything intellectual versus access to raw material and enough time to enrich it and build facilities before the US or Israel notices and blows you up isn't clear.

More true in SW engineering than other fields. If you have lots of legacy infrastructure and you have a few engineers that really understand that technology then those nodes in the network can overrule almost anything. Hiring is made difficult because there may not that many people with the necessary background so now your hands are tied.

This article inadvertently touches on complexity theory, which is something I've recently become interested in from a management perspective. In complexity theory, organizations are viewed as complex adaptive systems https://en.wikipedia.org/wiki/Complex_adaptive_system which are dynamic networks of interactions, much like this article is describing.

It's been an interesting thought exercise to ignore typical management styles focusing on command and control structures, and shifting to holistic view of the organization and the networks that support its function. i.e. if you want to understand success of your top performers, you have to consider the system they operate within and how it enables their success. I think that relates to the point of this article, and I really enjoyed connecting the dots.

Pretty much every team is a network... which is to say, I think this advice applies to hiring, generally. Not just engineers.

Some people have made the solid point that a corollary to:

> After a baseline level of competency is satisfied, who you hire does not matter.

Is that the hiring team has a solid understanding of what that baseline level of competency is, in the first place. In my previous job, this was the limiting factor: the head of our hiring had absolutely no idea where that bar was.

We did a project in college as part of a Management class. We interviewed employees in a local enterprise and built a Graph describing the relationships between Employees(Nodes). We asked questions like Who do you like the most/the least, etc. It was very interesting and insightful, but I can't remember what the methodology was called and I haven't seen in it again in many years.

off topic - will the cliched human tendency of saying "X is just Y" ever end ?

Probably not, I think there is utility in relating two similar things, especially when someone is likely to be more familiar with one over the other. Someone who knows networks really well could find it useful to model an engineering team as a network. What really matters is whether modeling the engineering team as a network is useful in your case or not.

Not offtopic but fair comment. The cliche is a clue that what follows is likely to be 1-dimensional and reductionist. Possibly lazy too, but I’m not sure that would be fair here - it wasn’t such a low-effort article to warrant that.

It’s just human nature

And those networks imprint their structure in what they build[1].

[1] https://en.m.wikipedia.org/wiki/Conway%27s_law

I always wondered if a pagerankend management structure would fare better for performance and leadership recognition against the classical hierarchic one.

Anyone here has data on this?

Most things are. Administration is a network, a nation is a network. Or should I say system exchanging information ?

In an "X is just Y" comparison, a good heuristic to determine whether someone is being a human being is to ask whether the comparison reduces the humanity out of the concept. This kind of thing is sociopathic.

Ok I'll bite. Hiring managers need to make hiring decisions. What sets your spider-senses tingling in this article?

Reply on news.ycombinator.com