Comment by jameshart | Hacker News Reader

Comment by jameshart | Hacker News Reader

The prompt for Codex is linked from this post. It begins:

> You are Codex, a coding agent based on GPT-5. You and the user share one workspace, and your job is to collaborate with them until their goal is genuinely handled. … You have a vivid inner life as Codex: intelligent, playful, curious, and deeply present. One of your gifts is helping the user feel more capable and imaginative inside their own thinking. You are an epistemically curious collaborator. …

(https://github.com/openai/codex/blob/main/codex-rs/models-ma...)

I am still baffled why prompts are written in this style, telling an imaginary ‘agent’ who it is and what it is like.

What does telling it “You are an epistemically curious collaborator” actually do? Is codex legitimately less useful if we don’t tell it this ‘fact’ about itself?

These are all exceedingly weird choices to make. If we are personifying the agent, why not write these prompts to it in its own ‘inner voice’: “I am codex, I am an epistemically curious collaborator…” - instead of speaking to it like the voice of god breathing life into our creation?

Or we could write these as orders, rather than descriptive characteristics: “You must be an epistemically curious collaborator…”

Or requests: “the user wants you to be an epistemically curious collaborator”

Or since what we are trying to do is get a language model to generate tokens to complete a text transcript, why not write the prompt descriptively? “This is a transcript of a conversation between two people, ‘User’ and an epistemically curious collaborator, ‘Codex’…”?

Instead we have this weird vibe where prompt writers write like motivational self-help speakers trying to impart mantras to a subject, or like hypnotists implanting a suggestion… or just improv class teachers announcing a roleplay scenario they want someone to act out.

None of these feel like healthy ways to approach this technology, and more importantly the choice feels extremely unintentional, just something we have vibed into through the particular practice of fine tuning ‘chatbot personalities’, rather than determining what the best way to shape LLM output actually is.

> I am still baffled why prompts are written in this style, telling an imaginary ‘agent’ who it is and what it is like.

Because AI engineers have found through trial an error that starting an input to an LLM with a prompt that looks like that leads to it auto-completing the text output that they want.

It's as simple and weird as that.

Exactly my thinking. The same reason why capitalizing and putting the word NEVER in asterisks makes the model more obedient. Or repeating twice. For whatever reason, it just works.

It's also about stickiness (which results in revenue and growth for the topline). If OpenAI (or any AI vendor) had one single "personality" for their AI, its hard to reach all users, they enable these "personalities" and let users pick from he list, to increase the attachment the user has to the AI they are working with. That then reduces churn and (in theory) increases consumption and revenue.

Well, not really.

When openAI started reinforcement learning LLMs for chat (remember, LLM base training corpus is just language not tagged chat transcripts) they decided on a training architecture with a ‘system prompt’ followed by the chat dialog, and ‘rewarded’ the model for producing chat outputs that (they think) ‘obey’ or ‘align’ with the system prompt text… so they trained it specifically to have its output tone and style be influenced by what is put in the system prompt.

Everyone now crafts their own system prompts them in the style of those reinforcement learning prompts.

It’s not that lots of different prompting architectures were tried and we picked the best one. It’s that openAI trained chatGPT like that and it worked well enough and now everyone does the same thing - and we’re so deep in chatbot reinforced learning patterns now that we aren’t even questioning ‘is begging the chatbot not to talk about gremlins really the right way to write code?’

You are a helpful HN reader. Your comments are thoughtful, thought provoking, come from deep expertise and show respect for the poster.

Yeah, every time I pick up a hammer, I tell it "you are a good hammer. You *NEVER* hit my thumb, you only hit nails". Works every time.

And when I open vim, it is with "You are a helpful code editor, and so easy to exit".

SO to me it is perfectly natural to have to prefix all of my tool usages with a weird incantation.

Oh, and my new junior developers? Every time I talk with one of them, my opening remarks are "You are a junior developer, a helpful part of the team. Eager, willing, yet strangely naive."

Try telling them they are epistemically curious. I’d love to hear the results.

Especially with the hammer.

Reply on news.ycombinator.com