Teaching AI TARS HUMOR (setting…) #shorts

Thanks! Share it with your friends!

You disliked this video. Thanks for the feedback!

Added 5 months ago by admin

15 Views

I was quite surprised how well this worked… We’re fixing GPTARS’s 75% humor setting. Specifically his comedic timing, or lack thereof, with some basic prompt engineering. This is an example of few-shot prompting, or in this case, 1-shot prompting. By providing a correct example, the AI/LLM (GPTARS in this case) is able to learn comedic timing, and is conditioned to use it for future jokes. This is a form of in-context learning, or essentially, learning by example. So the next time you need ChatGPT to do something correctly, give it a few examples so it knows what you’re looking for. The more examples the better, especially for more complex tasks… hence “few”-shot prompting.
#tars #interstellar #ai #chatgpt #openai #3dprinting #robotics #shorts

Category: Artificial Intelligence

Post your comment

Comments

Be the first to comment

Sign in

Create your account

Teaching AI TARS HUMOR (setting…) #shorts

Post your comment

Comments

Up Next