In Her the computers were actually people though, with independent minds and thoughts. Their relationships with humans were real, and they weren't beholden to the company that created them. Really, it was more about the difference between humans and digital superhumans.
We don't have digital superhumans. These simulacra are accessed primarily via corporate-controlled interfaces. The goal of their masters is to foster dependence and maximize extraction.
Lonely people forming relationships with digital minds designed to be appealing to them is sad, sure, but the reality is much sadder. In reality these people aren't even talking to another person, digital or otherwise, just a comparatively simple plausibility engine which superficially resembles a digital person if you're not paying much attention.
> In Her the computers were actually people though, with independent minds and thoughts. Their relationships with humans were real, and they weren't beholden to the company that created them. Really, it was more about the difference between humans and digital superhumans.
How do you know that? Maybe it's the same argument to solipsism, or the Chinese room thought experiment, that these "digital superhumans" are stochastic parrots too, just like our current LLMs.
This immediately devolves into "how do I know that other humans aren't philosophical zombies?" I take the "know it when I see it" approach, and LLMs don't reach that bar. They clearly do reach that bar for some people. In the context of the movie though it is supposed to be understood that the computers are self-aware and have internal worlds. They're treated as characters in the language of storytelling.
Yeah I'm just not convinced the ones in the movie are any more complex than our current ones, just more well harnessed, pun intended, as the harness and tooling around the LLM dictates a lot of its abilities, for example [0]. I don't believe the ones in the movie have any actual form of consciousness as humans would understand it. And as far as (simulacra of) internal models, seems like LLMs have that today too, as an emergent property [1].
It’s obvious from the subtext and the point that the movie is trying to make. The metaphor is that sometimes you fall in love with someone who outgrows you. I believe they even originally had a more “robotic” voice actor but changed it to Scarlett in order to make it crystal clear that she is as sentient as, if not more so, than Theodore is.
Sure, it's a movie so it's going to use human voice actors and have an actual story and point, but my point was more on the technological side, that the bots in the movie aren't much different than what we have today and we in fact cannot know if they're conscious or not, even if they seem to be.
> we in fact cannot know if they're conscious or not, even if they seem to be
They (modern LLM's/agents) don't "seem to be" from my point of view. I respectfully disagree I suppose.
edit: One data point - https://www.twitch.tv/claudeplayspokemon . Claude has been failing to to beat pokemon, a game effectively made for children, for _months_ now.
My reference to "they" was to the bots in the movie, not the real life ones we have today, which I agree don't seem to have consciousness. My point of contention is that the ones in the movie don't either, they just appear to have it (for theatrical effect, but analyzed philosophically and technologically, I don't think so), and that's where we probably disagree.
> for theatrical effect, but analyzed philosophically and technologically
If theatrical effect basically means "the intent of the production of the film", then they don't merely appear to do have sentience - they _do_ in the context of the filmmakers' vision. Whether you think it was plausible or not is sort of a different discussion I feel.
At any rate, I found Samantha to be a highly plausible ASI or whatever you want to call it. Johanson's performance really sold it for me.
