3 ms·

They are hard and there are many smart people working in the trades. I am not disagreeing with this premise. It is still difficult to outright deny the general

They are hard and there are many smart people working in the trades. I am not disagreeing with this premise. It is still difficult to outright deny the general intelligence gap between those who work in manual labor/trades jobs and those who work in STEM professions. Of course there are different types of intelligence. There are figures starting on page 80 in this document that support this claim: https://gwern.net/doc/iq/ses/2002-hauser.pdf https://gwern.net/doc/iq/ses/2002-hauser.pdf

strix_varius 3y ago
You're using data from 1944, based on fixed-pay government jobs, to support a claim that private sector software engineering is well-paid, in 2024, due to "IQ gating." The irony is hard to miss.
- atleastoptimal 3y ago
  1. it’s a good study because they gave the army intelligence test to everyone who came in, so little selection bias 2. a job doesn’t necessarily pay well just because it does require a high IQ. Scientists in a research lab, professorships, academic research all likely have more intelligent people on average yet don’t pay as well as software engineering. 3. The study still shows a routine and reliable difference in average IQ between those who work in manual jobs/trades and those who work in science/engineering Software engineering in the US is well paid, yes, partially due to economic effects and impact of software in and of itself, but it is, especially in FAANG, a profession with a very high concentration of people near the higher end of normal and above in the IQ spectrum, which leads to implicit selection criteria regarding intellectual capacities in employment.
  - nwiswell 3y ago
    #2 actually undercuts your argument about "IQ gating". Pay for software engineers is the product of supply and demand, and the absurd salaries clearly have much more to do with the demand side of the equation because of explosive growth in tech; if it really had to do with supply due to "IQ gating", then research scientists, professors, etc would make more. In any event there is more recent data (1992-4) that confirms that the differential is still present. This is maybe unpleasant to confront because it's self-congratulatory and has obvious class implications, but it's also pretty obviously true. See page 89 / figure 12: https://users.ssc.wisc.edu/~hauser/merit_01_081502_complete.pdf https://users.ssc.wisc.edu/~hauser/merit_01_081502_complete....

strix_varius 3y ago

You're using data from 1944, based on fixed-pay government jobs, to support a claim that private sector software engineering is well-paid, in 2024, due to "IQ gating." The irony is hard to miss.

atleastoptimal 3y ago
1. it’s a good study because they gave the army intelligence test to everyone who came in, so little selection bias 2. a job doesn’t necessarily pay well just because it does require a high IQ. Scientists in a research lab, professorships, academic research all likely have more intelligent people on average yet don’t pay as well as software engineering. 3. The study still shows a routine and reliable difference in average IQ between those who work in manual jobs/trades and those who work in science/engineering Software engineering in the US is well paid, yes, partially due to economic effects and impact of software in and of itself, but it is, especially in FAANG, a profession with a very high concentration of people near the higher end of normal and above in the IQ spectrum, which leads to implicit selection criteria regarding intellectual capacities in employment.
- nwiswell 3y ago
  #2 actually undercuts your argument about "IQ gating". Pay for software engineers is the product of supply and demand, and the absurd salaries clearly have much more to do with the demand side of the equation because of explosive growth in tech; if it really had to do with supply due to "IQ gating", then research scientists, professors, etc would make more. In any event there is more recent data (1992-4) that confirms that the differential is still present. This is maybe unpleasant to confront because it's self-congratulatory and has obvious class implications, but it's also pretty obviously true. See page 89 / figure 12: https://users.ssc.wisc.edu/~hauser/merit_01_081502_complete.pdf https://users.ssc.wisc.edu/~hauser/merit_01_081502_complete....