ChatGPT took the world by storm when it launched last year. It was seen as the next Google as people flocked over to Edge, Bing, and ChatGPT but more recently, concerns have begun to arise regarding whether ChatGPT is being nerfed on purpose. This might sound like a far-out conspiracy theory but there’s actually a surprising amount of truth to it. In fact, some researchers from Stanford and UC Berkeley have been studying the accuracy of ChatGPT over time and they’ve found some concerning declines in accuracy. In certain areas, ChatGPT’s accuracy fell by as much as 96% from 98% to just 2%. Some people have speculated that OpenAI is nerfing ChatGPT on purpose in order to force people to pay for the premium version. This might be part of the story but there’s more to it than just that. One of OpenAI’s biggest concerns with ChatGPT is limiting its use for nefarious purposes. It’s very possible that trying to nerf ChatGPT’s response to nefarious questions has affected its overall effectiveness. This video takes a look at the objective decline of ChatGPT and tries to identify some potential reasons for the decline.
Socials:
https://www.instagram.com/hariharan.jayakumar/
Discord Community:
https://discord.gg/SJUNWNt
Timestamps:
0:00 – An Objective Decline
2:35 – The Nerfing Argument
6:00 – The Woke Argument
9:22 – Our Fault?
Resources:
https://pastebin.com/3whh2WuD
