I've worked (and currently work) with a large number of people who are here on H1-B visas. A lot of them have been great people, exactly the kind of person who we should want to be immigrating here. Overall they're intelligent, hard working, and will be fully assimilated into American culture within a generation or two (at most) if they stay.
That said, none of them have been experts (much less unique experts). They're (usually) junior level developers that still require training to be fully productive. When the demand for IT exploded over the last 20 years, our departments doubled or tripled in size (maybe even more than that) to keep up. We needed the bodies. But that growth is flat now. AI might start driving it down. We don't need to keep that pipeline fully open anymore.
As a project manager specializing in new product development, I watched too many companies spend two or even three years recruiting for a "bleeding edge" highly specialized engineer, when they could have hired the most likely talent that just finished their BS and trained them up for a couple years in the specialization they needed. Of course, I spent nine years on active military duty, where the well known axiom was "we don't hire generals, we only hire lieutenants and train them", so I admit to already drinking that Kool-Aid. A lot of companies can't be convinced of the NPV of training; we need a better way to quantify the loss when not having the people you need on hand when a strong opportunity presents itself but you have to pass on it.
This is a good steelman of a crux of disagreement. I for one think that Graham is right, but I explicitly identify as an open borders globalist and see no reason why state power should, or even ethically can, be used to force companies to train up 'Muricans rather than productive people from the rest of the world, since the people from the rest of the world are no less valuable as human beings than 'Muricans and deserve no less freedom of movement and employment. I recognize this is an unpopular view and frankly doubt Graham himself fully shares it, but I think it is much more common among those who agree with him on this issue than among the general population.
One phrase I've been trying to work into conversations is 'we don't know how to do that'
So you say the most qualified person should get the job - great but how do we do that? All qualification testing is an imperfect approximation, a best guess at future performance under uncertainty
I've worked (and currently work) with a large number of people who are here on H1-B visas. A lot of them have been great people, exactly the kind of person who we should want to be immigrating here. Overall they're intelligent, hard working, and will be fully assimilated into American culture within a generation or two (at most) if they stay.
That said, none of them have been experts (much less unique experts). They're (usually) junior level developers that still require training to be fully productive. When the demand for IT exploded over the last 20 years, our departments doubled or tripled in size (maybe even more than that) to keep up. We needed the bodies. But that growth is flat now. AI might start driving it down. We don't need to keep that pipeline fully open anymore.
As a project manager specializing in new product development, I watched too many companies spend two or even three years recruiting for a "bleeding edge" highly specialized engineer, when they could have hired the most likely talent that just finished their BS and trained them up for a couple years in the specialization they needed. Of course, I spent nine years on active military duty, where the well known axiom was "we don't hire generals, we only hire lieutenants and train them", so I admit to already drinking that Kool-Aid. A lot of companies can't be convinced of the NPV of training; we need a better way to quantify the loss when not having the people you need on hand when a strong opportunity presents itself but you have to pass on it.
This is a good steelman of a crux of disagreement. I for one think that Graham is right, but I explicitly identify as an open borders globalist and see no reason why state power should, or even ethically can, be used to force companies to train up 'Muricans rather than productive people from the rest of the world, since the people from the rest of the world are no less valuable as human beings than 'Muricans and deserve no less freedom of movement and employment. I recognize this is an unpopular view and frankly doubt Graham himself fully shares it, but I think it is much more common among those who agree with him on this issue than among the general population.
One phrase I've been trying to work into conversations is 'we don't know how to do that'
So you say the most qualified person should get the job - great but how do we do that? All qualification testing is an imperfect approximation, a best guess at future performance under uncertainty