When I read the original question, I thought that yes - he was essentially a bully, but he probably didn't think of himself as one. But that smiley face - he must have felt at least somewhat like a bully just then.

It's not a race to the bottom. It's a race to the point where Supply&Demand curves intersect.

P.S. if it was a race to the bottom, consider that only 1.5% of the labor force works at minimum wage.

You completely misunderstood. In an idealized free labor market scenario, the wages of the scarce specialist employees would go up to the point that Apple or Google could barely make a profit. The race to the bottom is on profits, not wages. Just like small businesses typically compete on sales price until they only have a trickle of a profit.

To avoid this, companies use various tricks to fix wages such that they don't have to pay even the highest valued employees more than a fraction of the value they actually bring to the company, and thus can get away with huge profits.

To illustrate this with an extreme scenario, assume that all of Apple's work relied on a single engineer's designs. If Google and Apple were to compete on who can hire that engineer, Apple would be willing to pay almost all the way up to their entire revenue to that single engineer, as in this contrived scenario, if they lose him, they lose all of their revenue. So, Google and Apple would be racing to the bottom, offering ever more extravagant salaries to this magical engineer until they couldn't pay him and make a profit at all anymore.

Gross profit is the limit to reference in this thought experiment, not revenue.

The thing is the idealised situation never actually exists. Let's say a developer builds a product that brings in $10m a year. How much do you pay them? You need to factor in that actually 10 designers, artists and product managers also contributed to the design and functioning of the product. Then there's the 5 people in IT support, 2 in QA, 3 in customer support, the office staff and security, the finance people, payroll, marketing, distribution, etc. Maybe it actually costs $7m a year to run all of that.

So you pay them $3m a year? Now there's zero profit, so nothing to invest in expanding the team, improving your services services, hiring more developers, growing the businesses, funding new projects for that developer and all the other employees to work on. Oh, by the way you owe $20m to your investors that put down the capital to get the business off the ground in the first place. How are you going to pay them back? Also, maybe you could have hired another developer and they might have done just as good a job, which is also true of everyone else in the business.

Finally, if you do up the pay of the developer say $1m a year, what happens if the product gets eclipsed in the market or the developer moves on to another product nowhere near as profitable. Now they're working on something bringing in a lot less, but they're still costing you a million a year.

Man, I hope some six-figure Bolsheviks read this comment and think about it. It's really not as simple as people like to think to try and pay you workers millions

Even Facebook has had to lay people off when the market demand for $500k genius coders cooled a little recently

FAANG salaries and RSUs are pretty much the perfect compromise between labor and capital IMO. The only way I'd change things would be to move toward fewer hours per week, possibly with widespread fixed-bid consulting and health benefits not tied to employment

In microeconomics, one of the consequences of the long-term equilibrium is that no company in the industry has any profit. Sure, it's just a theory that doesn't capture everything but it's often referenced for such cases.

I think it's actually worse than that for the companies, in particular it can easily be a race to negative profits for all companies involved. For example, imagine two competing bridges over a river, where the total revenue available is $100. Each bridge is built of 50 bricks, which collect salary and are free to work for whichever bridge gives them the better offer. The maximum salary that's sustainable if the bridges make a profit is $1 per brick, but it's worth offering vastly more than $1 to "bribe" any particular brick from the other bridge to defect since doing so causes your only competitor to fall down. Similarly, your competitor should rationally be willing to counter-offer up to their entire profit to any single brick to prevent that from happening. Subject to some assumptions about how negotiations work, this can result in both bridges being really far in the red, since the total value of the bridge/team is worth less than the sum of the costs of each individual brick/member departing. The decision theory concept of a "dollar auction" has some similar pathological behaviors.

Obviously, software engineering teams aren't as dramatically interdependent as a bridge where the departure of any one member causes the entire structure to collapse, but they also don't live in the perfect balance where the value of the whole exactly equals the sum of the marginal value/cost of each member. As a result, there's no one number for the "value" of an individual engineer that both represents their negotiated wage and is viable for the company to pay without going bankrupt.