100ProofTollBooth
Dumber than a man, but faster than a dog.
No bio...
User ID: 2039
100ProofTollBooth
Dumber than a man, but faster than a dog.
2hr ago
Diversity is our Strength. Us being whites
At the top of Marginal Revolution today: "How Cultural Diversity Drives Innovation"
I'm a tech development and "innovation" nerd. There's a small, but growing, especially in recent years, online commmunity of people who read organizational histories of places like Bell Labs and the original Lockheed Skunkwords to try and figure out the best ways to do real tech development. Not academic science projects and not VC backed bullshit which is mostly business model innovation (that even more often fails).
You don't have to read the whole study. The abstract itself is either a hilarious self-own or and even more hilarious playing-dumb post.
We show that innovation in U.S. counties from 1850 to 1940 was propelled by shifts in the local social structure, as captured using the diversity of surnames. Leveraging quasi-random variation in counties’ surnames—stemming from the interplay between historical fluctuations in immigration and local factors that attract immigrants—we find that more diverse social structures increased both the quantity and quality of patents, likely because they spurred interactions among individuals with different skills and perspectives. The results suggest that the free flow of information between diverse minds drives innovation and contributed to the emergence of the U.S. as a global innovation hub.
1850 to 1940. Bruh.
This paper shows that having big time diversity - you know, mixing all those crazy Poles, Irish, French, Germans, English, Welsh, Czech, Slovak, Greek, hell even a few Italians and Spanish in there - was a massive reason the USA was such a technologically innovative place!
The HBDers are going to love this one.
Side note on the hard tech angle: patent issuance used to be a decent enough and standardized enough measure for "innovation." Since the rise of legalism post WW2, however, it's so much more noisy now that it's questionable if it remains a valid "fungible currency" for studying innovation and tech development.
100ProofTollBooth
Dumber than a man, but faster than a dog.
2hr ago
Are there any pre-modern wars where a soldier could be sent out to the front line, and then 2-3 years later in the war, find himself in almost the exact same spot, despite regular bursts of fighting?
No, but this has more to do with feeding and supplying an army than anything else. The modal soldier in pre-modern warfare might spent 2 - 3 years more or less walking in a giant, slow circle, almost starving to death every day. And then actually starving to death.
100ProofTollBooth
Dumber than a man, but faster than a dog.
2hr ago
Forgive me for the "akshually" style comment, but this isn't entirely true.
No one has come up with a VERY CHEAP effective countermeasure yet. The ones that do work are 1) expensive and 2) Horded by the US/ISR/China and (maybe) a few other countries because nobody wants to show off their cool-new-shit in Ukraine. We want to save it for when it - yikes - actually matters.
Much like the human element of the Ukraine war, the drone element is mostly one of attrition and competing supply lines. At one point, 10,000 drones were falling out of the skies over Ukraine per month because of effective and cheap countermeasures. The tactical wheel turns, however, and both sides elevated their drone-counterdrone game.
100ProofTollBooth
Dumber than a man, but faster than a dog.
2hr ago
In a word, The Holodomor.
Now, don't worry, I'm not some Ukraine agent apologist here. I'm just trying to directly answer the question of "What are the Ukrainian people afraid of, being conquered by Russia?" You can absolutely boil Ukraine v Russia down to Red Tribe vs Blue Tribe. The Ukrainians aren't thinking about the future, they're constantly enraged by the past. The "Politics of Resentment" isn't an invention of 21st American politics - it's the de facto arrangement of most human conflict. To many in Ukraine, allowing a Russian takeover is the equivalent of letting all of the people who killed all of your family members move in to your house. It's pretty easy to get fatalist and irrational to prevent that. "I would rather die than ...." Yeah, well.
It's quite close to 0% planned and 100% spontaneous.
Kuhn's The Structure of Scientific Revolutions is probably still the best framework for how human knowledge (science and so, downstream, technology) develops. The long and short of it is that lots of happy accidents often build upon each other. Planning innovation is almost an oxymoron.
The problem then becomes, how do we 'cultivate the garden', so to speak, to make happy accidents more commonplace? Or to shorten the distance between related but unknown nodes that are working on the same problems? The University System and the various Bell Labs / PARC / DARPA orgs of the mid 20th century seem to have done this well. Both had different failure modes which roughly follow red and blue tribe cleavages.
The University System lost to ideological capture but also, more generally, a total remove from practical problems. Instead of a bunch of really smart professors working with Corporations, the Navy, or whomever or an actual problem, "pure" research began to win out. You'd get esoteric improvements in something like photonics that was utterly untenable in a production setting because the supply chain for the super rare materials didn't exist or the apparatus involved couldn't function outside of a clean lab.
The Bell Labs etc. failed because corporations stopped funding them. There's a debate as to why. Some simply gesture at "grrr greedy capitalists" which has never been a satisfying answer for me. The better answer, though still not "a-ha!" level in my mind is that actually novel and meaningful research is getting harder and taking longer. So, while a corporation may not need its R&D department to come up with something new every quarter, it's harder to not want to cut their budget after 10 or 20 years of nothing new. Furthermore, there's a pretty good argument to be made that corporations shouldn't be trying to shoot-the-moon with totally novel ideas but, rather, really be solving the "last mile" problem of new technology - how to sustain it, scale it, and then make it by degrees cheaper and cheaper. The middle ground that's evolving is something like Focus Research Organizations.
The final players - DARPA and other FFRDCs (Federally Funded Research and Development) kind of kept the spirit alive longer. DARPA has a very specific operating model that nowhere else in government replicates. But they fell victim to GWOT funding strategy - let's make everything about terrorists instead of focusing on, I don't know, time travel and teleportation. The FFRDCs became some of the most egregious leeches of Federal R&D welfare dollars. MITRE is quite literally make work jobs for PhDs. If you can endure living in a Kafka novel every day, you can make $200k per year and enjoy Tysons Corner traffic for your commute.
The real "oh, we fucking suck" moment was GPT-2 in late 2022. Almost every other major American technology development since WW2 could be traced back to some sort of federal, academic, or corporate R&D lab. That the Attention Is All You Need paper came out from a some ML engineers at google fucking around was, in my mind, kind of the tombstone on the "trad" R&D ecosystem.
More options
Context Copy link