You seem to keep suggesting that boundaries have to be perfect for categories to exist. This is not the case. White colour and dark colour exist as categories yet are blurry. Yet no sane person suggests colours don’t exist.
There’s no sharp boundary between a hill and a mountain. There’s no sharp boundary between dark and white colours. There’s no sharp boundary between male and female either at the edge cases when you consider intersex. Language has its limitations and it’s obvious to most of us and we don’t have to play semantic juggling each time.
Does the way people use race as a word have biological basis? Absolutely yes. For example you take people who self identify as black and white and try to find whether their DNA can predict their self identified race. You would agree that it can predict to a good extent I hope.
On a side note: I think you have fallen into this popular rabbit hole where you expect sharp boundaries for commonly used words/categories. Such sharp boundaries never exist and every normal person from a child to an adult knows and works around this. This kind of rhetoric is popular with some extremist academic types to signal their in group presence. Tip: if you want to solve racism, don’t get into debates on esoteric definitions of race and pretend race itself doesn’t exist. Normal people aren’t convinced.
What even is biological race? Race as used sociologically has a biological basis. And normal people think that race is real, the same way colour is real.
You have said this
> Races don't exist on any biological level
And it is completely untrue. Sociological race does exist at a biological level. DNA of blacks and whites are different.
Just like wavelength emitted by black and white colours are different.
The survey asked whether race is real - of course race is. Just like colour is real.
White and black are not biological races but words used to describe real clusters of people, even if its lossy. In the entire conversation, you have made a complete strawman, as if people are talking about some made up definition of biological race you have come up with and then claim that "ha that made up definition is not real!"
Skin color and (facial) bone structure for example. Like it or not, but humans have adapted to living in different climates and have different ancestral lines. Take a picture of an average black person and edit to white skin, it will still not look like a white person.
As the commenter above you already said, this is not a bad thing. People on _hacker_news should be able to understand that being able to define the ==/!= operators on something does not automatically mean that you can define < and >.
It is more obvious with some and a bit more nuanced with others, but in general yes. Just like you can typically see the flat/wide nose in Africans and Asians but not Europeans. Again, there is no implication of one being "better" than the other here, but don't pretend it's impossible to categorize people like this.