TLDR: Recently ChatGPT just started wildly lying, inventing bullshit and disregarding my instructions. The root cause seems to be that OpenAI has instructed it to reduce use of the search tool (you can see references to this in the chain of thought). I imagine this is to conserve resources, but without some factual context to go on ChatGPT hallucinates like hell.

But there is a better way: Use the command line version of your preferred AI tool. Watch NetworkChuck's video for details! These are available for Gemini (free), ChatGPT and Claude (both of which require a standard subscription).

Why is it better? You can force the AI to remember rules and context for any given project, agent workflow, or output style. How? You can create a directory for each project, and add a file describing the project, background, rules and output or analysis style you want. When you start up the AI from that directory, it will read the background files and immediately know the project goals and expected behaviour without you tell it. And it will use that information as context for every query. So it is much less likely to drift off base, and much more likely to follow your instructions, and you can customise it for different roles such as editing articles, working on coding projects and so on.

As described in the video, you can use your AI directories and context files with any AI you like. And if you create a private Github repository for your project you have all the history as well.