These two tweets are interesting:
"One lesson that GPT-3 might be teaching us: the tell-tale signature artifact of simulation is not the spiky glitch, it's the smooth, shallow, facile surface. It's not the errors, it's a certain kind of boring flawlessness, that's what we should be on the lookout for." @flantz
"Being able to distinguish between layers and categories of meaning is going to be a highly valuable skill for a while. Good news for snobs, weirdos, artists, nerds, and librarians." @flantz
Thoughts that occur:
There's a kind of writing that is actively aspiring to a smooth, shallow, facile surface, a certain kind of boring flawlessness. The writing that most corporations do. That's who GPT-3 will get to first. Not the novelists and the poets but the corporate copy-writers. They came for the writers of car brochures, but I wasn't a writer of car brochures, so I said nothing.
The translation is already roboticised. Perhaps the writing is too.
And presumably GPT-3 is already being deployed by SEO agencies and, equally presumably, Google Search spooks are already trying to simulate the skills of snobs, artists, nerds and librarians.
Hopefully that means that a premium will get attached to quirky non-smooth writing. I have a colleague who uses 'dang' a lot in her writing. Unusual on this side of the Atlantic. I think that'll get through a copy captcha.
It also reminds me of something William Gibson writes in a Paul Smith book. It's something like: if you want to know what's good about what your country makes look at what the Japanese and British import. (I can't find the actual quote). And his reasoning is something like: Japanese and British cultures are so hierarchical and stratified, in such long-lasting and subtle ways, that they've become incredibly good at discerning status from tiny signals.
Which makes me wonder what the 'looking at their shoes' of writing is going to be.
Dang.