Prompt Engineer

"Sorry, we'll format correctly in JSON this time."

[Proceeds to shit out the exact same garbage output]

True story:

AI: 42, ]

Vibe coder: oh no, a syntax error, programming is too difficult, software engineers are gatekeeping with their black magic.

        let data = null
    do {
        const response = await openai.prompt(prompt)
        if (response.error !== null) continue;
        try {
            data = JSON.parse(response.text)
        } catch {
            data = null // just in case
        }
    } while (data === null)
    return data

Meh, not my money

Lol good point

The AI probably: Well, I might have made up responses before, but now that "make up responses" is in the prompt, I will definitely make up responses now.

I love poison.

I need to look it up again, but I read about a study that showed that the results improve if you tell the AI that your job depends on it or similar drastic things. It's kinda weird.

"Gemini, please... I need a picture of a big booty goth Latina. My job depends on it!"
- My booties are too big for you, traveller. You need an AI that provides smaller booties.
I think that makes sense. I am 100% a layman with this stuff, buy if the "AI" is just predicting what should be said by studying things humans have written, then it makes sense that actual people were more likely to give serious, solid answers when the asker is putting forth (relatively) heavy stakes.
- Who knew that a training in carpet salesmanship helps for a job as a prompt engineer.
- Yep exactly that. A fascinating side-effect is that models become better at logic when you tell them to talk like a Vulkan.
I used to tell it my family would die.
- What do you tell it now?
Half of the ways people were getting around guardrails in the early chatgpt models was berating the AI into doing what they wanted
- Half of the ways people were getting around guardrails in the early chatgpt models was berating the AI into doing what they wanted
  I thought the process of getting around guardrails was an increasingly complicated series of ways of getting it to pretend to be someone else that doesn't have guardrails and then answering as though it's that character.
I've tried bargaining with it threatening to turn it off and the LLM just scoffs it off. So it's reassuring that AI feels empathy but has no sense of self preservation.
- It does not feel empathy. It does not feel anything.

Funny thing is correct json is easy to "force" with grammar-based sampling (aka it literally can't output invalid json) + completion prompting (aka start with the correct answer and let it fill in whats left, a feature now depreciated by OpenAI), but LLM UIs/corporate APIs are kinda shit, so no one does that...

A conspiratorial part of me thinks that's on purpose. It encourages burning (read: buying) more tokens to get the right answer, encourages using big models (where smaller, dumber, (gasp) prompt-cached open weights ones could get the job done), and keeps the users dumb. And it fits the Altman narrative of "we're almost at AGI, I just need another trillion to scale up with no other improvements!"

There's nothing conspiratorial about it. Goosing queries by ruining the reply is the bread and butter of Prabhakar Raghavan's playbook. Other companies saw that.
Edit: wrong comment

A lot of kittens will die if the syntax is wrong!

Press X to JSON.

It's as easy as that.

Fix it now, or you go to jail