Hello,
Our company is seeking an AI expert/prompt engineer who can either develop a new or fine-tune an existing AI model for us. We are an AI-based paraphrasing startup company that aims to enhance our paraphrasing capabilities. We currently bypass modern AI detectors 70-80% of the time, but we want to achieve more consistent results. If you are interested in working with us, please let us know. Our company currently uses GPT-3 and GPT-4 to do such things, but we're looking to explore and expand to other models to improve our product.
It would be best to create a prompt and/or fine-tuned GPT to bypass AI detectors such as GPTZero, TurnItIn, and more. After conducting our testing, we've found that modifying temperature and penalty settings tends to bypass such systems; however, it's not always effective compared to our biggest competitors.
We will need the prompt to be protected from prompt injection attacks and prevent our prompt from being revealed or the model from doing tasks other than paraphrasing.
This job is already closed and no longer accepting applicants, sorry.
