I'd start working from https://beta.openai.com/playground/p/default-micro-horror?mo...
Write a paragraph describing the battle between an African Lion and a bison.
The battle between an African Lion and a bison was fierce and intense. The bison, weighing nearly a ton, attempted to use its immense size and strength to overpower the lion. Despite his size, the lion was no match for the bison's strength and agility. The lion used its superior speed, agility and razor sharp claws to try to inflict damage on the bison. Although the lion was able to land multiple scratches on the bison with its claws, it was not enough to take down the massive animal. After a long and grueling fight, the lion ultimately retreated in defeat as the bison emerged victorious from this epic battle of nature.
https://imgur.com/a/z6il6IZGPT3 is stateless. You give it the prompt and it responds back. Context doesn't carry between requests unless you (the user) pass it along. That can make "chat" sessions with GPT (not ChatGPT) expensive as passing along the past context of the chat session consumes a lot of tokens as it gets longer and longer.
It is possible they are using GPT itself to summarize the past transcript and pass that in as context. So your “remember X” would probably get included in that summary.
That said, I have not tried to tell it to remember X and then exceeded the 8192 token contest window.
Source: https://twitter.com/goodside/status/1598882343586238464
https://beta.openai.com/playground/p/default-chat?model=text-davinci-003
https://beta.openai.com/playground/p/default-friend-chat?model=text-davinci-003
https://beta.openai.com/playground/p/default-marv-sarcastic-chat?model=text-davinci-003
you can see a chat "session".The issue is that to maintain the state, you need to maintain the history of the conversation.
For example, the friend chat starts out at 28 request tokens. I submit it, get a response back, and it's at 45 tokens. I add in a bit of my own commentary... and it's 59 tokens. I hit submit and its 80 tokens.... and so on.
That session wasn't particularly insightful, but you get the idea.
The issue is that I did three requests with 28 + 59 + 88 tokens. As I keep going, this can add up. To maintain that context, my previous history (and the chat response) remains in the prompt for the next message growing arithmetically.
---
(edit)
Working from the chat instead of friend chat... a bit more interesting conversation.
Note however, that at this point the next request will be at least 267 tokens. We're still talking fractions of a penny ($0.002 / 1k tokens) - but we are talking money that I can look at rather than even smaller fractions. And as noted - it grows each time you continue with the conversation.
(I understand lots of people want the full tailored experience and don't want to write 10 lines of Python, but my main point is that this is easy for someone to build today. I think you don't see too many companies doing this because there is exactly zero moat.)
I just whipped this up as a demo:
It's very basic but demonstrates the basic idea if you want to go get an OpenAI API key and give it a shot. You can give it more sophisticated prompts to get different sorts of answers.