Anyone that remembers the reaction when Sydney from Microsoft or more recently Maya from Sesame losing their respective 'personality' can easily see how product managers are going to have to start paying attention to the emotional impact of changing or shutting down models.
If you want an LLM to retain the same default personality, you pretty much have to use an open weights model. That's the only way to be sure it wouldn't be deprecated or updated without your knowledge.
Consider the implementation: There's document with "User: Open the pod bay doors, HAL" followed by an incomplete "HAL-9000: ", and the LLM is spun up to suggest what would "fit" to round out the document. Non-LLM code parses out HAL-9000's line and "performs" it at you across an internet connection.
Whatever answer you get, that "personality" is mostly from how the document(s) described HAL-9000 and similar characters, as opposed to a self-insert by the ego-less name-less algorithm that makes documents longer.
For example, keep the same model, but change the early document (prompt) from stuff like "AcmeBot is a kind and helpful machine" to "AcmeBot revels in human suffering."
The document or whatever you'd like to call it is only one part of the story.
I brought up prompts as a convenient way to demonstrate that a magic-trick is being performed, not because prompts are the only way for the magician to run into trouble with the illusion. It's sneaky, since it's a trick homo narrans play on ourselves all the time.
> The document or whatever you'd like to call it is only one part of the story.
Everybody knows that the weights matter. That's why we get stories where the sky is generally blue instead of magenta.
That's separate from the distinction between the mind (if any) of an LLM-author versus the mind (firmly fictional, even if possibly related) that we impute when seeing the output (narrated or acted) of a particular character.
I think LLMs are amazing technology but we’re in for really weird times as people become attached to these things.
I’m less worried about the specific complaints about model deprecation, which can be ‘solved’ for those people by not deprecating the models (obviously costs the AI firms). I’m more worried about AI-induced psychosis.
An analogy I saw recently that I liked: when a cat sees a laser pointer, it is a fun thing to chase. For dogs it is sometimes similar and sometimes it completely breaks the dog’s brain and the dog is never the same again. I feel like AI for us may be more like laser pointers for dogs, and some among us are just not prepared to handle these kinds of AI interactions in a healthy way.
It's Reddit, what were you expecting?