That's just it: American tech companies aren't staffed by Americans anymore. Liang Wenfang, the founder of DeepSeek, made the point that he values developing domestic Chinese talent over importing foreign experts. It seems to have worked.
"It seems to have worked" I am sure he's a great manager, but whats really working for him is crazy high number of graduates and PHDs in China who are unemployed or underemployed
That would sound a lot like the languishing native STEM talent in the US if it weren't also for the fact that most of the DeepSeek team doesn't have PhDs.
I should clarify: the US STEM workforce has historically been very white and very male. That's the body of talent that has been languishing, and the data proves it.
> That's a complete non sequitur. Why would it make sense for new hires to be overwhelmingly disproportionately oversampled from minority groups according to current demographics?
Because that is not what this stat is measuring. They are doing (demos of new employees - demos of retiring employees), not just demos of new employees. Hence why it is a misleading statistic. They word it very carefully to not say 94% of new hires are minorities.
source: the "The Analysis" section of your own source or if you needed it stated more explicitly [0]:
> Before judging whether that’s impressive or excessive or some other adjective, it’s helpful to know what the available pool of new workers looked like. Or, more precisely, what the pool of new workers minus the pool of departing workers looked like. Net change is what we’re able to see.
i'm sure this will have precisely 0 impact on your worldview though
since your reply died - your absolutely wrong about massive oversampling, you can get 95% with basic assumptions about 60% white, 40% minority (matching actual proportions in the population - and actually an underestimate when you consider that new hires are young) and the retiring fortune 500 population (70-90% white).
