This is a story about a human who has developed a ~relationship with an LLM and is therefore way out of my usual zone—the “AI girlfriend” articles always feel like rubbernecking a car crash to me—but Kashmir Hill is such a great writer and uses this fundamentally sad and tawdry situation to tease out some genuinely haunting dynamics. (The context window section made my soul lurch.)
nytimes.com/2025/01/15/technol…
(archive.ph also has a copy if you want one)
Anne Deschaine
in reply to Erin Kissane • • •Elena ``of Valhalla''
in reply to Anne Deschaine • •@Anne Deschaine @Erin Kissane TBF, in Star Trek the AI beings were actually sentient people, and proving that sentience was often the point of the episode.
The current “AI” things are just text (or image, or video) generators based on randomness and probabilistic analysis of existing works, and we know that there isn't any hint of sentience in that.
I'd blame the abuse of the term AI for marketing purposes.
like this
Erin Kissane, Anne Deschaine e a wandering happenstance like this.
Anne Deschaine
in reply to Elena ``of Valhalla'' • • •@valhalla True, but it wouldn't be hard to find the same sorts of points to argue now just reading that article! 😏 Even in those episodes, the ultimate conclusion often relied on one person's willingness to take the next step. AI aside, we're still questioning what/who is sentient.
But anyway, mostly it makes me think about the AIs in Trek and wonder how started. Every Trek-like thing introduced today seems to have the most horrifying origin/growth story, and that is damn depressing.
Elena ``of Valhalla'' likes this.