I think it's a bit in between. FWIW, at 54 I've never failed a whiteboard interview, including at my current FAANG employer, so this is most certainly not sour

I think it's a bit in between. FWIW, at 54 I've never failed a whiteboard interview, including at my current FAANG employer, so this is most certainly not sour grapes. While not deliberately ageist, I think whiteboard interviews work out a bit that way. What can be tested in such an interview? Only a sampling of domain knowledge. It will tend to be a sample weighted toward the particular problems and algorithms that will be super-fresh in the minds of people still in or fresh out of school (usually because the interviewers themselves are). To someone older but still well versed in that domain, those particular details might have been crowded out by a thousand other things learned since. They might seem less familiar because of changes in language, notation, or idiom. Those same differences will also affect how the result is "graded" even if the candidate solves the problem quickly and well. Whiteboard interviews also fail to measure other things such as ability to select algorithms or higher-level approaches, people or organizational skills, industry knowledge, or a developed "instinct" about what symptoms suggest what problems in the relevant kinds of code. As more weight is given to whiteboard skills, less weight is given to literally everything else. As I said, I don't think any of this intentionally disadvantages older workers, but it can have that effect without intention. It has never hurt me personally, but I have plenty of peers who I know beyond doubt could code rings around the people who interviewed them. They just couldn't dance that particular dance well enough in that moment and got rejected. That's a loss for the (potential) employer as well as for them.