Exactly! I have lost count of how many times people on HN rant about quality of "Indian software engineers" when they are paying so less that they get bottom of the barrel. And they make quick stereotype out of that experience without any introspection.

I've thought about the "worker shortages" that are a common theme in tech, I think a big part of the issue is that American's income has been flat-ish for the last 20 years: https://en.wikipedia.org/wiki/File:Historical_median_persona...

So when there are specific jobs that have scarcity and salaries are raised, many managers literally have no experience with that and don't know how to hire in a competitive market.

I've been in the banking industry in Asia, where the compensation is well above the average local.

The average quality of the Indian engineers (in banking), is unfortunately pretty abysmal.

The problem, I think, is their bias towards to hiring their own. Banking IT can become absolutely dominated by Indians, to the point where Indians can staff > 90% of IT positions in countries where they are a minority. The average quality of the Indians in there are pretty bad.

The difference can be seen clearly in companies that enforce some kind of diversity quota. The average quality of the Indians in there is noticeably higher by a wide margin. These companies are unfortunately a minority.

That's a component and a big component of it is the non-technical managers and wannabe founders that are working with the indian software engineerings and failing to articulate anything in anyway for any engineer to accurately deliver upon.

That's part of the software engineers job too is it not? They're the hired expert.

When I was consulting I spent half my being a product development/manager person because I needed to be.

This skill is likely what enabled you to do consulting. Understanding and critiquing what your building is a really important part of being an engineer, but it's relatively rare - which is a shame.

So much this.

A reductive view that isn’t quite correct. In a healthy job market there should be a smooth range of work in a field from lower value and lower pay up to higher value and higher pay. An uneven distribution means there are companies who struggle to justify the salary demands of an overqualified person because the work or the business model supporting the work doesn’t benefit from the surplus of value that employee brings. There’s nothing wrong with being that kind of company. In time the median skill level of Indian engineers will increase and solve this.

A smooth distribution for ML engineers doesn’t exist anywhere in the world, imo.

Is ML more of a team sport? With specialists in operations, data cleansing data, configuring models, etc.

We're offering 90 percentile compensation for Bangalore, we have excellent engineers in our offices there already. I'm just complaining about the bimodal nature of ML hiring in general. General software development is a bit more linear.