LLMs and all such products tend to converge: on styles, on forms, on patterns; the “sameness” isn’t incidental, it’s actually core to what these technologies do, how they work; they will always have this quality; this quality is their modus operandi, really! There is no world in which LLMs write interesting things.

Humans care about what other humans say in part because of the content and in larger part because of who says it / when / why / etc. The % of each varies by context, but techies who think it’s “mostly” about the content have it backwards.* If the person I admire the most on earth sent me an LLM-written pre-fab message, it would mean as much to me as a campaign email that starts with “Hello, friend” or whatever. Obviously! Obviously! I don’t care if the LLM “trained on their own work,” either. No one will, after some novelty wears off! And this is all obvious!

No one wants this except LinkedIn. The OP hates it; the “commenters” may like it, in some way, but since the OP and others hate it, they’re losing status / beclowning themselves; and they probably don’t like it much / get anything “real” from it. But LinkedIn sees various (proxy) metrics move and must feel good about it, even though those proxies are so disconnected from any user’s experience that they practically constitute a subversion of the platform’s real goals.

Anyway: bless all the fools out there.

-Mills, designer at Substack

People need to be more thoughtful building products on top of LLMs. The fact that they generate text is not the point.