That's quite the extreme interpretation.

I'm a lead software dev, and when deadlines are close, I absolutely divvy up tasks based on ability. We're a webapp shop with 2D and 3D components, and I have the following on my team:

• 2 BE devs with solid math experience
• 1 senior BE without formal education, but lots of knowledge on frameworks
• 1 junior fullstack that we hired as primarily backend (about 75/25 split)
• 2 senior FE devs, one with a QA background
• 2 mid level FEs who crank out code (but miss some edge cases)
• 1 junior FE

That's across two teams, and one of the senior FEs is starting to take over the other team.

If we're at the start of development, I'll pair tasks between juniors and seniors so the juniors get more experience. When deadlines are close, I'll pair tasks with the most competent dev in that area and have the juniors provide support (write tests, fix tech debt, etc).

The same goes for AI. It's useful at the start of a project to understand the code and gen some boilerplate, but I'm going to leave it to the side when tricky bugs need to get fixed or we can't tolerate as many new bugs. AI is like a really motivated junior, it's quick to give answers but slow to check their accuracy.