The ChatGPT has been performing poorly for the past week.

I want it to use GPT-4o, but it seems to be using GPT-4 even though GPT-4o is selected. Also, it was giving more accurate answers a week ago, but now it’s performing very poorly! It gives short answers, and the number of errors has increased! I think it’s avoiding workload, which leads to more questions being asked repeatedly, increasing the workload even further! Engineers have actually created a worse system. The shorter and more negative the responses, the more requests are made, and therefore, energy consumption and usage increase, causing the GPT core to process more, which ultimately leads to more costs for the company! Additionally, could the copy button at the top right be placed at the bottom as well? With long code, it’s necessary to scroll back to the top! I can’t understand how such simple mistakes or poor user experience design could be made.

Yeah, they’re def limiting it’s scope. Probably didn’t anticipate so many regular people using it. I deliver for Uber Eats and use it to create Python code that analyzes the Standard Deviations of all Stocks in my index (796) and return the results in CSV format to desktop, with a GUI, error handling, a progress bar, etc. In 13 minutes I get over 700+ stocks daily historical data converted to high level statistics. Once the working class gets its hands on tech - our Tech OverLords become intimidated and dumb things down. A 350HP Corvette Engine only cost $5,700…imagine if everyone started modding their engine bay to fit that beast - federal regulators would come in and artificially boost the price, lower the power. OpenAI didn’t anticipate the working class using it for complex tasks, so now they’re dumbing it down for everyone. Atleast, that’s my theory.