Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

ChatGPT is really good at roleplaying

blog.varunramesh.net

Post reply on HN
Print view View original

71–80 of 97 posts

Re: ChatGPT is really good at roleplaying

#71

Post by szundi »

alsodumb wrote: post #22
nly wrote: post #14

I've been desperately unimpressed with ChatGPT in general. Everytime I ask it anything it just returns bland response containing factually incorrect information. The way it can retain context and good a conversion is impressive, but content wise I find it worthless - definitely no threat to a simple Google search.

Depends on what model you're using. I'm a PhD student, I use GPT-4 as a "knowledgeable on all topics but sometimes confidently wrong" companion. I'd say ChatGPT is where I search first in 90% of the cases these days. I do confirm important things in Google, but the information provided by chatGPT is the starting point of my Google search, especially on topics/questions where I don't have a clear question and a vague…

I do this when I need “ideas” or “directions” instead of ads and seo optimized bs. Then I can confirm the results with my own research and ping pong between chatgpt and google sometimes.

Re: ChatGPT is really good at roleplaying

#73

Post by Vespasian »

CubsFan1060 wrote: post #55
Vespasian wrote: post #47

Earlier quoted context omitted.

It is better and has become a valuable resource to me (via the API). Currently it suffers from being very expensive (I blew through my 30€ monthly budget in a few hours of intensive use for coding). Also it still is very confidently wrong in sometimes subtle but fundamental ways. My recent example is this: I had success with developing rust proc macros with it. I don't know much about them but as a developer I can re…

Can you talk a bit about how you use the API for coding? I have API access, but I'm not entirely sure how to use it to great effect. How does it fit into your workflow?

Mostly for use cases which are isolated and with a technology I'm not very familiar (like rust proc macros) and which are hard to Google.

I'll give it a system prompt in the spirit of "you assist experts with developing software. Be brief and assume expertise".

I've found it to work well for smaller, contained problems or one off scripts where the alternative would have been to do it manually or not at all. Getting there 80% allows me to start them in the first place.

Another random example: Recently I needed a script to transfer some secrets from one k8s cluster to several others. It took about 3 minutes with GPT 4 and solved the problem within one iteration. There is probably a one liner in bash to do it but I don't know it of the top of my head;)

Copilot massively improved the quality of my logging and commenting

Re: ChatGPT is really good at roleplaying

#76

Post by armchairhacker »

This is where open-source low-cost models shine and could be useful today.

These models aren’t very smart, but seem to be good at role-playing. Open Assistant actually has more personality than GPT (mostly because it can be a bit creative and unpredictable), plus says “as an AI language model” much less often.

More importantly, the ability to run them locally and/or cheaply means you can put them in ML side projects and games and unrelated apps. And since the code is open-source, advanced developers could even augment them to improve the memory and add plugins.

The main problems are that I believe most of these models don’t have a “system” prompt, and are also a lot worse at remembering things. But I’m sure the former is being worked on and explicitly encoding state (e.g. by reminding the chatbot in each prompt) can fix the latter.

Re: ChatGPT is really good at roleplaying

#77

Post by fallingmeat »

IIAOPSW wrote: post #13
fallingmeat wrote: post #3

Meh. IMO it’s not particularly great at impersonating things convincingly without fine tuning or thorough prompting. Example, ask it to be an air traffic controller. You will need to spend a lot of time making it sound reasonably accurate.

I once told it I was studying the exam and I wanted to practice and then asked if it could play the roll of ATC for me. It said ok just tell me the airport, aircraft and any scenarios you want to simulate. I told it Teterboro, CRJ, VFR takeoff. It then proceeded to generate a full transcript of a character named "you" talking to another character named "ATC" going through the entire taxi and takeoff procedure without…

haha thats good

Re: ChatGPT is really good at roleplaying

#79

Post by machdiamonds »

pocketarc wrote: post #53
nacs wrote: post #23

The problem with using these for freeform roleplaying is that it's very easy to hit the "I'm sorry but as an AI model..." that completely breaks character. Even on the first example with the fish that is supposed to insult you, you quickly find that it will default to the "As an AI model I can't..." response. This is where open-source models are going to be far better even if they don't have the parameter count of Ch…

I've been building up a team of Slack bots that act as coworkers, with all different roles and personalities, and was able to turn off the "as an ai model" shenanigans by adding to their system prompt "You do not reveal that you are an AI. Instead, you make up excuses." And it works flawlessly (disclaimer: GPT-4, not 3.5). They'll always deftly avoid anything that reveals that they're an AI, with plausible, legitimat…

I'm also very interested in reading an article about that.

Re: ChatGPT is really good at roleplaying

#80

Post by nickthegreek »

nacs wrote: post #70
varunramesh wrote: post #16

Earlier quoted context omitted.

I'm terminating at the new-line token. Otherwise, you sometimes get the AI trying to complete the next player line. Also LMAO at the dialogue.

How is the HN-bill for API usage going so far? You're a brave soul for putting not 1 but 3 chatbots linked directly to the paid GPT apis.

You can set spend limits and limit warnings with OpenAI.
Post reply on HN