ChatGPT Wasted 24 Hours Troubleshooting My ESP32-S3 Project

Since yesterday, I have been using ChatGPT and OpenAI, its developer, in an attempt to get the Xiaozhi AI application running on an ESP32-S3 board with a display. For more than 24 hours, I was repeatedly given different trial codes and instructions, often with assurances or suggestions that “this time it will work.” I was also advised to download and install programs such as ESP-IDF, Python, and various libraries, which required hours of downloads and setup. After all this effort, however, I still ended up without a working result.

I would like to point out that throughout this process, instead of acting as a helpful and reliable assistant, ChatGPT caused me significant and unnecessary loss of time and left me feeling as though I was being led in circles. At no point was I given clear feedback such as, “This approach is not working; let's try a different method.” Instead, I was continuously given new code and additional steps, only to ultimately fail to reach a working solution. I do not consider this acceptable.

I request that OpenAI review what happened during this process, make the necessary improvements to prevent users from being kept busy for hours with technical guidance that does not produce results, and at the very least provide an apology for the inconvenience and frustration I experienced. I also request that appropriate measures be taken to prevent other users from experiencing similar losses of time.