ChatGPT is rendering other people's chat history in the sidebar

Post by Spivak » Mon, Mar 20, 2023, 5:13 PM UTC

jetrink wrote: post #35

croddin wrote: post #29

Earlier quoted context omitted.

Robert Tables strikes again!

I'm surprised we're making this mistake again. Isn't it possible to structure the chats as something more than a flat stream of text and insert unrepresentable tokens after the tokenization step?

Post by SketchySeaBeast » Mon, Mar 20, 2023, 5:17 PM UTC

chpatrick wrote: post #3

I always find it surreal when this super science fiction tech is possible in our time but basic stuff like this can't be implemented correctly.

Post by chatmasta » Mon, Mar 20, 2023, 5:19 PM UTC

behnamoh wrote: post #41

doctoboggan wrote: post #25

Last time something like this was posted, someone from OpenAI said those weren't other peoples chat history, they were randomly generated chat titles. (which is what happens when you prompt GPT with a blank line)

How can this happen in 2023? More importantly, how can a well-funded firm like OpenAI allow such lousy database bug to occur in their system? For all the hatred towards Google, I'm sure at least they wouldn't have this kind of problems once they open their (much more powerful) LLM.

Post by wongarsu » Mon, Mar 20, 2023, 5:24 PM UTC

Spivak wrote: post #51

jetrink wrote: post #35

Earlier quoted context omitted.

I'm surprised we're making this mistake again. Isn't it possible to structure the chats as something more than a flat stream of text and insert unrepresentable tokens after the tokenization step?

Yes and no. You can represent chats as structured data but you have to leave the airlock eventually and render it to flat text because the model only knows "complete the next word" and that in-band structure is context for the model itself to understand the flow of messages. So the model is seeing a series of text and using that to figure out who's talking but if you convince the model to output then when it goes to…

Post by addisonl » Mon, Mar 20, 2023, 5:27 PM UTC

random_cynic wrote: post #54

I like how people on Reddit cannot distinguish between ChatGPT itself and its web front end. I thought Reddit is programmer friendly. Has anyone checked if the services using the API have been affected or not?