Comment by satvikpendem
13 days ago
How is this specific to 4o? This can happen with any model. See how people acted after Character.AI essentially removed their AI "partners" after a server reset. They actually used DeepSeek before which didn't have the same limitations as American models, especially being open weight means you can fine tune it to be as lovey dovey as your heart desires.
From the subreddit I linked in another comment, there did seem to be some "magic" that 4o had for these kinds of "relationships". I'm not sure how much of it is placebo, but there does seem to be a strong preference in that user group.
4o was very sycophantic so was very willing to play along with and validate the users roleplay. OpenAI even noticed enough to talk about it in a blog: https://openai.com/index/sycophancy-in-gpt-4o/
> OpenAI even noticed enough to talk about it in a blog
That's one way of interpreting things...
6 replies →
It probably is placebo. Character AI for example used DeepSeek and I'm sure many grew attachments to that model. Ultimately though I don't even get it, models lose context very quickly so it's hard to have long running conversations with them, as well as talking very sycophanticly to you. I guess this is fixed due to implementing a good harness and memories, which is what these companies did I assume.
> as well as talking very sycophanticly to you.
That's apparently a feature to a significant amount of people..
One version of 4o was so sycophantic that it had to be rolled back, so there is some evidence that 4o specifically has a problem with this.
https://openai.com/index/sycophancy-in-gpt-4o/
After 4o they put in more safeguard reactions to the user attempting the kind of (lets be generous here) romantic roleplay that got a lot of people really invested in their AI "friends/partners".
ex: https://www.reddit.com/r/MyBoyfriendIsAI/comments/1qx3jux/wh...
I think 4o was more than just unusually sycophantic. It “understood people” better and had a kind of writerly creativity.
I used it to help brainstorm and troubleshoot fiction: character motivations, arcs, personality, etc. And it was truly useful for that purpose. 4.5 was also good at this, but none of the other models I’ve tried.
Of course this particular strength is dangerous in the hands of lonely unstable people and I think it’s dangerous to just have something like that openly out there. This really shows that we need a safe way to deploy models with dangerous specializations.
I'm of the persuasion that if people need help, it's better to get them that help instead of nerfing the tools for everyone.
I agree with you, but safeguarding the vulnerable while preserving access for the fit is something we as societies know how to do, if we try.
> Of course this particular strength is dangerous in the hands of lonely unstable people and I think it’s dangerous to just have something like that openly out there.
Thank you for saying this
People are not happy with this because 4o, at least from what I've heard, seems to be much more willing to go down the relationship/friend path than 5.2 and Claude and the like.
[dead]
I can't believe they would stoop so low as this kind of character assassination.
It’s great marketing though