We may just be working under different definitions. Are you claiming that when I type things into, say, Hacker News and hit reply, the words you read aren't the words I wrote?

Or are you asserting the "person" of the words in the computer is not the same person I am behind the keyboard?

I'd argue that the latter is the disproven hypothesis. It turns out people who said awful things online were actually awful people; they may not show it as often in public, but they weren't different human beings. Broadly speaking, they believed the things they said and tended to act on them in real life.

Laughing off things on the computer as not real was how at least one shooting went unchecked.