That is not necessarily the truth; the Copilot has digressed and has become the least intelligent AI model on the market. Regardless of the prompt and even when receiving exact prompts from better AI models, it still fails.
Why has the quality of this program deteriorated to being equivalent to a spell check and calculator? It has become increasingly worse, with the system often just agreeing with me or fabricating stories when I request articles on specific topics. It lists off made-up names and situations, and when I demand citations, it apologizes, claiming it didn't realize I wanted true examples, despite it being stated in the input.
In attempting to troubleshoot work or computer programs, technical support, etc., it repeats the same incorrect answer and does so repeatedly. When I ask for a new answer, it changes the title, like "Alternative Solutions," but provides the same response.
Last summer, the system was decent, but developers have continued to downgrade it. ChatGPT fixed two days' worth of problems on both of my PCs (problems created by Copilot) within an hour. I feel deserving of a refund because this service is not worth $20 a month; it should be a $3.99 per month phone app.
Why is Microsoft unable to provide a decent AI model? If the AI makes up answers and uses outdated data, what value does it offer?