There’s been an extensive marketing campaign to convince people that LLMs are intelligent. I wouldn’t call someone a subhuman for assuming there is some truth to that.

Of those that understand what an LLM is, I think you can divide them into two groups, the honest, and the dishonest. Honest people see no use in a bullshit generator, a lying machine. They see it as a perversion of technology. Dishonest people have no such objection. They might even truly see intelligence in the machine, as its outputs don’t differ substantially from their own. If you view language as a means to get what you want, rather than a means to convey factual information, then lying is acceptable, desirable, intelligent. It would be difficult for such a person to differentiate between coherent but meaningless bullshit, and a machine with agency making false statements to pursue its own goals.