Actually it's not an absurd take at all. The absurd take is that we "should not bend language around ignorance."
That's precisely how language changes over time. Language is not a strict set of rules. It's based on understanding and consensus, so sometimes things that are "wrong" do end up being accepted.
I suggest this as a great introduction into what languages are and how they evolve over time https://www.amazon.com/Language-Families-of-World-audiobook/...
Most mistakes remain mistakes, and do not become part of the language. The idea that mistakes generally get accepted as correct is simply untrue, which is what you are implying.
I am sure people will make the mistake, as they sometimes do today. But it is a mistake, and will likely be recognised as one.
It is likely that the language gets more cemented by automatic spelling and grammatical correction, including using AI. For example, there are a number of grammatical and spelling changes that have been cemented by American spelling/grammar checking programmes ie. by MS Word.
I am not a native speaker, but the two words do not sound even remotely the same.
How does this mistake happen so often? Can you explain people's thought process a bit? Is it just: "Something something 'seg...' ... ah I know, I will simply use another random word that starts with the same 3 letters and doesn't make sense in this phrase!"?
Also this is the first time I see it.