I would guess there’s an upward slope of young contributors that reduce significantly at post-grad and early career ages (e.g. early-mid-twenties) followed by another upward gradient on a 5-10 year delay that peaks in the late 30s then falls somewhat linearly up to 60s.
The median age of the younger/learning cohort might be 19 and falling. The median age of the established developer cohort might be mid-40s and climbing.
Stingrays = Kleenex of IMSI catchers