It's also a little misleading to compare the numbers that way. It usually takes a lot longer to become a psychiatrist in the first place. Your cumulative earnings in the first ten years after graduating high school might be higher as a plumber; in fact, I suspect most people who become doctors actually lose net worth throughout those ten years due to student loans. They just don't count towards the median income for a psychiatrist because they spend much more time not being a psychiatrist.
Once the plumber and psychiatrist are both in a position where they're running their own business or practice, their income is more of a conscious choice than it is for someone who's working for an employer. Self-employed people can and do sometimes choose to earn less money in exchange for more free time, and it's a lot easier to do that when you don't have six figures of debt at the start of your career.
Plumber median pay is $54,000, while 10% make more than $94,000.
Psychiatrist median pay is $220,000, while 10% make beyond the wage cap for BLS data. https://www.bls.gov/oes/current/oes291066.htm
This one sticks out to me from a job-market POV.
Plumbers are already on par with psychiatrists in terms of pay. Sure, the work is hard, but from a supply-demand POV, the markets are saying that they do work that is of equal value (all this varies highly, but stick with me here).
Most repair work takes a LOT of experience until the employee is anywhere near proficient. Farmers will tell you that they learned the repair side in very stressful situations and that there is always something that breaks in new, and more expensive, ways each and every time. Repairing even 'simple' machines like a tractor takes a lot of experience.
So, when you add in a computer that intentionally 'breaks' the machine, and then you have the regular machine that will break, and then you have all the linkages that make the computer control the machine (all of which can break), you are requiring that the repair-person be VERY skilled at troubleshooting and repair.
Then, you eliminate most of the ways that people learn to troubleshoot and repair these machines (via these computers), and you are left with even fewer people that will be able to fix the tractors. Think something more like a bio-tech/MRI company's field-service engineers. Those people get paid bank to get sent out to malfunctioning machines and they are trained out the wazoo.
From the tractor manufacturer's perspective, yes, making machines that work for the manufacturer and not the customer, is going to really work out well for this quarter's stock price. But down the road 7 years? I think this is a poison pill.