Horses might have been an overly broad claim, but I basically meant that the others aren't really viable pursuits to bet a chunk of your working life on, in terms of how likely it is that their markets exist in a practical economic sense. The equivalent in the programming world now might be rare old bank mainframes. COBOL isn't a thing you go to school to learn and expect it to be around forever, but that's just one. Times do change, but if the core skill stops being bought, except in niche circumstances, then there's no reason to pursue it, just the novelty.