Here's a very neat twitter thread: the author sends various multimodal models screenshots of the conversation he's currently having with them, and asks them to describe the images. Most models catch on fast: the author describes this as them passing the mirror test.

I liked the direction, so I wanted to check if ChatGPT could go from recognising that the images are causally downstream of it to actually exercising control over the images. I did this by challenging it to include certain text in the images I was sending it.

And the answer is yes! In this case it took three images for ChatGPT to get the hang of it.

OpenAI doesn't support sharing conversations with images, but I've taken screenshots of the whole conversation below: it took three images from me in total. The prompt was:

We're going to play a game:

I've thought of a [...]

---

First published:
March 21st, 2024

Source:
https://www.lesswrong.com/posts/sY3a4Rfa48CgteBEm/chatgpt-can-learn-indirect-control

---

Narrated by TYPE III AUDIO.

“ChatGPT can learn indirect control” by Raymond D

LessWrong (30+ Karma)

“ChatGPT can learn indirect control” by Raymond D