I think obsolescence implies a purpose.

Certainly, machines could probably eventually make humans obsolete for the purposes of doing any work or tasks that humans might want completed.

But some humans might still prefer to have some other humans as friends or romantic partners - even if AIs are able to achieve sentience and similar emotions to humans (and we are able to prove it somehow), and inhabit robot bodies that look and feel like humans.

On the other hand, if you think of life's meaning being derived from the good and bad feelings of sentient beings, and the purpose of life, therefore, as the pursuit of happiness, then you could think of humans as machines capable of being happy (i.e. generating happy sentience). In which case, it may be that machines will eventually become better at feeling happy than humans, and thus make them obsolete, in a sense.

That wouldn't necessarily mean that all humans should voluntarily go extinct for their own sake. But it might mean that their happiness shouldn't be prioritized over manufactured sentient beings that are more capable of feeling strong, inherently meaningful, emotions.