The older intellectual — and the younger one. They might have been talking about Nietzsche. Or Rajasthani taxi cab drivers. Sept. 2018
In the process of discussing Microsoft’s stack ranking system, that certainly led to massive productivity loss at Microsoft, and maybe even to the collapse of several market dominance technologies, my co-conspirator Ryan Martens sent me the picture below, from Tomasz Tunguz’ blog
(Credit Tomasz Tunguz — Venture Capitalist at RedPoint)
tagged to an article about Conway’s Law that so much of my work is constructed around. Tomasz’ interpretation of Conway’s Law is considerably less nuanced than my own, and not as well-developed. Still, I learned a couple of things, most notably that the Harvard Business Review turned down Melvin Conway’s attempt to publish his theory in 1967. Conway went on to get his Ph.D. from my B.S. alma mater, Case Western Reserve University.
Tunguz’ org. charts for product development are, as followers of this blog know, incomplete, because they don’t account for:
a.) the directionality of those little arrows — what’s the direction of information flow?
b.) the information coherence embodied by those little arrows – how much are people paying attention on either side of any given node to what is being said, and how are their v-Memes interpreting this?
c.) the flexibility of re-direction — how much does the org. chart discriminate from people going outside the org. chart to form their own relationships?
Every organization has an org. chart that kinda looks like one of the above (I love the Microsoft one with the little guns!) that is produced by the C-suite. But every organization also has an actual org. chart, that will dictate the real structure of product design. Such real org. charts are also time-dependent. There may be a time when everything in a smallish organization may get reviewed by, say, a master mechanical designer for conformance, as well as an overview of unnoticed synergies, and that doesn’t show up on ‘who reports to whom.’ The expanded version of Conway’s Law we talk about is really a thing, and a fundamental law of the universe.
I did love (well, maybe found stimulating) the fact that the HBR turned down Conway’s paper. It really helps my argument that lower v-Meme publications, no matter how sophisticated, are gonna shoot down more evolved thinking with regularity.
On another note, Tomasz is obviously an ‘ahead of the curve’ thinker, and I’m going to recommend reading his blog. He’s also a venture capitalist, and an author himself. His latest book is Winning with Data: Transform Your Culture, Empower Your People, and Shape the Future (Wiley). Might have to pick it up and give it a look. I’ve also written Tomasz and will see if he writes back. He’s got questions. We’ve got answers.
Oh — and if you need a Sunday Reading, I highly recommend the obvious short post referenced in this blog, as well as the Vanity Fair article at the beginning of this piece. This whole notion of competition uber alles as the path to success is effectively destroyed in this excellent piece. Jack Welch’s (former CEO of General Electric) philosophies, that are really the roots of all this, carried to extreme have really wrecked American business, and created the mess with shareholder capitalism that’s currently destroying our country.
And you might review this as well on the Intermediate Corollary, which puts that Knowledge thing in the middle of Social and Design structure. Not having an independent article on the Intermediate Corollary is a deep flaw of the blog. I think I’ve written about it a ton in all my academic papers (that no one reads, BTW) that I’d forgotten about being more explicit here, where people actually come. As of this writing, I’m up to about 33K views…
2 thoughts on “I’m Not the Only Person Interested in Conway’s Law”
I wonder if the ‘older intellectual’ is M Conway.
(Since i didn’t know who he is, I looked a little bit at his web page, and also blog posts about teaching high school in Massachusetts to disadvantaged students; his remarks seemed ‘spot on’. Context is very important. One size does not fit all—one curriculum isn’t appropriate or optimal for every school, and one curriculum is not appropriate or optimal for every student in a school. This is one reason people are trying homeschooling or creating alternative schools.
I guess using one curricula seems efficient—just mass produce it (and see what you get). Economists study ‘fair division ‘ problems –how you divide resources among different subgroups of a big group , and then among individuals in each subgroup, so its fair–using game theory in general.
Their answers to me seem pretty abstract (find some Nash or Pareto optimal equilbria in some cases) so I don’t know how applicable they are to the rel world.)
I like those corporate pictures or graphs. I’ve studied a bit of ‘chemical graph theory’ and polymerization theory—where you represent chemicals by graphs. The difficult math problems come in when you have graphs with loops , rather than trees.
Sometimes people don’t even deal with the loop cases, which are most relevant, because the math is too difficult, or they find some way to make a correction to what they could compute. .
Google, FB and Apple graphs have loops, and are all succesful in various degrees. Amazon is as well, but doesnt have any loops, so it seems to be an anamoly. (Amazon books does have alot of reader comment on its site—a kind of open peer review, so perhaps there are some implicit loops.)
Microsoft to me (as an almost complete outsider to business and silicon valley, tho i know one person with a job at FB) is still succesful, or at least they have famous people on staff (M Freedman, who won a Fields medal for solution of 4D Poincare’s conjecture in topology (Smale won one for D>4, and Perelman won for D=3, which is seen to have closed the case. Perelman also refused to take the $1million prize because he was unhappy with his experience in math community.
When i went to college the day before i was to get my degree, i told them i didn’t want it. My parents came up there and said we spent all this money so you have to take the degree–for them, i did..
They say a degree (especially the kind i got) is worth alot of money, nice job, etc. Not for me.)
I like that Microsoft graph—did they hire the NRA for their management team? The one for Oracle looks good too. It looks like its partitioned into two main colored subgraphs–the famous ‘graph partitioning problem’. Looks like they have the optimal partition—3 managers, HR people or lawyers per one engineer who installs the light bulb.