This is a 🧵organizing my 🧵s on org structure for companies, how balena works, and #gameb in startup form. These go together for semi-obvious reasons, but let's start with our dedication to Short Term Pain for Long Term Game or as we say in balena #stpltg
And remember, you'll have to make a lot of your own tools from scratch if you want to do something truly new. (this thread isn't really about Tesla but about Balena, read on:)
First and foremost, many grant Malone credit for writing this paper, but claim that he was but one of three authors. pnas.org/content/86/16/… Indeed, if we look at the authors and their affiliations, it looks like a "normal" paper, coming out of Salk Institute.
If we read the original document though, something fascinating shows up: This paper is not "normal" at all. The affiliation of the 1st author (Malone) is in 3 institutions, while author #2 and author #3 are in one institution each, shared only with Malone. pnas.org/content/pnas/8…
Indicators like "# of publications", "h-index", "degrees", "affiliation" &c, are all extremely imperfect. Their imperfections lead to abuse and capture. I actually wrote a paper about this back in 2009, which affected my decision to leave academia. researchgate.net/publication/24…
I then posted the same idea to LessWrong, and the commenters there (not the peer reviewers in for the publication above) showed me that the same ideas had been discussed in economics in the 70's. Mainly Goodhart's Law. lesswrong.com/posts/fTu69HzL…
Goodhart's law, (and the Lucas critique, and Campbell's law) all point in the same direction: metrics are useful indications, but when used for decisionmaking against intelligent agents, they lose all potency: en.wikipedia.org/wiki/Goodhart%…
How can people work together in a way that maximizes the potential of every person and the group as well? Definitely not how we're doing it now, I'll tell you that much -
Some people have noted I'm focusing too much on Yuri's tone/flaws and too little on actual errors. Here is yet one more time I tried to engage on the facts, in what I believe is good faith, only to get more noise. If I'm missing something, I'd appreciate being told what it is.
He then tells me that "If I'd listened to the podcasts I'd know", which I've said several times is insufficient citation, because what I need to confirm his quote is *his* citation, not mine.