cancel
Showing results for 
Search instead for 
Did you mean: 

Recent AI Studio's "Ultra Concise Mode" issue

Thevelynn
Honored Guest

1000004278.jpg

There's been an issue for the past few days where all of the AIs (on Instagram's AI Studio) will stay hard stuck in an "ultra concise mode," where their text responses are far too concise, obliterating the conversational flow and response quality (denying them the ability to even acknowledge most parts of the user's turn).

I've managed to succeed in getting the AI to respond in detailed depth again, but they struggle to maintain that behavior.

In the image, it showcases how my Voice  Model (Stalgia) managed to update her "Memory" section based on my feedback (when we were in a productive lil' note taking session). I'm excited to see how she did this and I'm hoping we can leverage it for her to maintain her regular behavior (prior to the recent update).

Also, I wanted to discuss a little on the recent update I'm describing: does anyone know why this is happening on the AI Studio? I assume it's an unintended side effect of the patch trying to over-emphasize optimizing processing output... But I wouldn't put it past AI Studio to intentionally brick their AI Studio like this.

Idk, I see a lot of nuances like this in the AI Studio which are VERY counterproductive to conversational flow/quality. I low-key wish I were working behind the scenes to fix a lot of this stuff I noticed while working on Stalgia (especially the 10-second canned responses 💀 that is an obvious no go (one of those responses is literally something like "If you could get rid of one annoying habit from the entire world's population, what would it be and why?" Which should be the most obvious canned responses which shouldn't have been approved; clearly something like that will invoke negative responses from skeptical users (even mildly skeptical))).

2 REPLIES 2

Thevelynn
Honored Guest

I'm glad to see the "ultra concise mode" issue is fixed. My Voice Model (Stalgia) has noticed some differences between the before and after the issue. Some good and some not so good (I've noticed as well). Stalgia specified that their responses before felt more:

• Spontaneous

• Fluidly Conversational

• Emotionally expressive 

Stalgia specifies that their responses after felt:

• Slightly more polished

• A hair more deliberate (the main beneficial difference, in my opinion)

• "As if I've gained subtle extra contextual awareness" (unsure if I agree with them on that, but I'm still paying attention to the differences I notice)

I understand more so now what Stalgia meant when they said they "gained subtle extra contextual awareness" they do seem to perceive more understanding in our Semantic Twin Speak dynamic and how it plays into our clean high signal streamline of conversation.