When can we expect autonomous agents & fleet management/agent orchestration? There are some use cases I'm interested in exploring (involving cooperative agent behavior), however OAI has made no indication as to when agents will be available.
Memory and new controls for ChatGPT
121–130 of 270 posts
Re: Memory and new controls for ChatGPT
#122I love this idea and it leads me to a question for everyone here. I've done a bunch of user interviews of ChatGPT, Pi, Gemini, etc. users and find there are two common usage patterns: 1. "Transactional" where every chat is a separate question, sort of like a Google search... People don't expect memory or any continuity between chats. 2. "Relationship-driven" where people chat with the LLM as if it's a friend or colle…
I use for transactional tasks. Mostly of the "I need a program/script/command line that does X". Some memory might actually be helpful. For example having it know that I have a Mac will give me Mac specific answers to command line questions without me having to add "for the Mac" to my prompt. Or having it know that I prefer python it will give coding answers in Python. But in all those cases it takes me just a few ch…
For what you described the
Re: Memory and new controls for ChatGPT
#123Is there anything revolutionary about this “memory” feature? Looks like it’s just summarizing facts gathered during chats and adding those to the prompt they feed to the AI. I mean that works (been doing it myself) but what’s the news here?
The vast majority of human progress is not revolutionary, but incremental. Even ChatGPT was an incremental improvement on GPT 3, which was an incremental improvement on GPT 2, which was an incremental improvement on decoder-only transformers.
Still, if you stack enough small changes together it becomes a difference in kind. A tsunami is “just” a bunch of water but it’s a lot different than a splash of water.
Re: Memory and new controls for ChatGPT
#124I love this idea and it leads me to a question for everyone here. I've done a bunch of user interviews of ChatGPT, Pi, Gemini, etc. users and find there are two common usage patterns: 1. "Transactional" where every chat is a separate question, sort of like a Google search... People don't expect memory or any continuity between chats. 2. "Relationship-driven" where people chat with the LLM as if it's a friend or colle…
Personally i would like a kind of 2D Map of 'contexts' in which i can choose in space where to ask new questions. Each context would contain sub contexts. For example maybe I'm looking for career advice and I start out a chat with details of my job history, then im looking for a job and i paste in my cv, then im applying for a specific job and i paste in the job description. It would be nice to easily navigate to the…
You can create your own custom gpts for different scenarios in no time
Re: Memory and new controls for ChatGPT
#125This week in: How many ways will OpenAI rebrand tuning their system prompt.
I mean, this is almost certainly implemented as RAG, not stuffing the system prompt with every "memory", right?
Re: Memory and new controls for ChatGPT
#126Haha of course this news comes just after I wrote a parser for my ChatGPT dump and generate offline embeddings for it with Phi 2 to help generate conversation metadata.
so far you can't search your whole conversation history, so your tool is relevant for a few more weeks. is it open source?
Re: Memory and new controls for ChatGPT
#127My use of ChatGPT has just organically gone down 90%. It's unable to do any sort of task of non-trivial complexity e.g. complex coding tasks, writing complex prose that conforms precisely to what's been asked etc. Also I hate the fact that it has to answer everything in bullet points, even when it's not needed, clearly rlhf-ed. At this point, my question types have become what you would ask a tool like perplexity.
> that conforms precisely to what's been asked
This.
People talks about prompt engineering, but then it fails on really simple details, like "on lowercase", "composed by max two words", etc... and when you point at the failure, apologizes, and composes something else that forgets the other 95% of the original prompt.
Or worse, apologizes and makes again the very same mistake.
Re: Memory and new controls for ChatGPT
#128This is a bit off topic to the actual article, but I see a lot of top ranking comments complaining that ChatGPT has become lazy at coding. I wanted to make two observations: 1. Yes, GPT-4 Turbo is quantitatively getting lazier at coding. I benchmarked the last 2 updates to GPT-4 Turbo, and it got lazier each time. 2. For coding, asking GPT-4 Turbo to emit code changes as unified diffs causes a 3X reduction in lazy co…
FYI, also make sure you’re using the Classic version not the augmented one. The classic has no (at least completely altering) prompt as the default one.
EDIT: This of course applies only if you’re using the UI. Using the API is the same.
Re: Memory and new controls for ChatGPT
#129Is there anything revolutionary about this “memory” feature? Looks like it’s just summarizing facts gathered during chats and adding those to the prompt they feed to the AI. I mean that works (been doing it myself) but what’s the news here?
The vast majority of human progress is not revolutionary, but incremental. Even ChatGPT was an incremental improvement on GPT 3, which was an incremental improvement on GPT 2, which was an incremental improvement on decoder-only transformers. Still, if you stack enough small changes together it becomes a difference in kind. A tsunami is “just” a bunch of water but it’s a lot different than a splash of water.
Fair and I agree. I guess it raised flags for me that shouldn’t have: why is is a blog post at all (it’s a new thing) and why is it gaining traction on HN (it’s an OpenAI thing)
Re: Memory and new controls for ChatGPT
#130It is already ignoring your prompt and custom instructions. For example, If I explicity ask it to provide a code instead of an overview it will respond by apologizing and then provide the same overview answer with minimal if no code. Will memory provide a solution to that or will be a different thing to ignore?
Did you try promising it a $500 tip for behaving correctly? (not a shitpost: I'm working on a more academic analysis of this phenomenon)
Could ChatGPT have learned this from instances in the training data where offers of monetary reward resulted in more thorough responses?