Using Whisper.CPP to (Cheat an Interview or) Listen with Live Audio and Question ChatGPT

First, The Repo and Explanation

While this program works, I wouldn't even consider trying to use something like this for a number of reasons. I just wanted to see if I could do it. Especially when I saw that there options out there to do this sort of thing. Very proffessional and expensive solutions. I actually think something like this would ruin an interview. You would lose your train of thought. But given that.. here it is:

Quick Summary

The application listens for 2 key phrases. One starts the prompt that will be sent to ChatGPT and the other let's the app know when your finished with your prompt. It continues to listen but also sends your prompt off to ChatGPT. Displaying the response back in the terminal. Could this help someone cheat on some interview question? Probably. Will it get someone a job? Higgghhllly doubtful.

When you interview you need to respond quickly and can't afford to lose your train of thought considering if some response from an LLM is right or not. In a longer lasting coding challenge.. yeah it could help I bet.

An image of me generating this blog post from audio

Is right here

Using Whisper.CPP to Listen Live and Question ChatGPT - Possibly Cheating an Interview

In today's digital age, the job search process has become more competitive than ever before. As a recent job seeker, I found myself overwhelmed with the sheer number of applicants vying for the same positions. It seemed like no matter how hard I tried, I just couldn't stand out from the crowd.

That's when I stumbled upon services that offered to use ChatGPT or other large language models to help cheat with interviews. These services claimed to provide candidates with the perfect responses to common interview questions, giving them a leg up on the competition. However, there was a catch - they charged exorbitant amounts of money for their services.

I couldn't believe that people were willing to pay such high fees for what essentially amounted to cheating in an interview. I thought to myself, "I could just build something like this myself." And so, I set out to create a solution that would level the playing field for job seekers everywhere.

That's how I came up with the idea for a repository that would provide job seekers with the tools they need to ace their interviews without resorting to cheating. By leveraging the power of ChatGPT and other language models ethically and responsibly, I hoped to empower candidates to showcase their skills and experience in the best possible light.

As I dove into the world of natural language processing and machine learning, I realized just how powerful these technologies could be in helping job seekers navigate the often daunting interview process. With the right tools and resources at their disposal, I knew that anyone could improve their chances of landing their dream job.

So, if you find yourself in a similar situation, feeling overwhelmed by the job search process and tempted by the allure of cheat services, remember that there are ethical ways to leverage technology to your advantage. By taking the time to hone your skills, craft compelling responses, and showcase your unique strengths, you can stand out from the competition and land the job of your dreams. And who knows, maybe one day you'll be the one creating repositories to help others succeed in their job search journey.

Actualllllllyyyyy.... The Above is Written by the Application I Created

I just dictated:

I need you to write a blog post about me searching for a job and coming across services that use ChatGPT or other large language models to help cheat with your interview. They charged crazy amounts of money. and I thought I could just build something like this myself which is why I created this repository I believe

I thought it would be ironic to have it write the blog post about this too as I don't really care to put anymore effort into this.

Give it a try though. You just download a model from Hugging Face, and then run the application with the model path as an argument.

I even provided 2 scripts to make and run the application in the repository.

./makeStreamGPT.sh # will download the model and build the application

# And to run

./runStreamGPT.sh Should start listening to your microphone and outputting the text to the console.

Good luck out there.. The world is a scary place right now for me as a software engineer. Job security wise. I shouldn't be wasting my time with this and be focusing on LeetCode or finding a challenging security/GoLang/Rust related problem to solve...