OK, it works for me, but in 10 minutes of testing against Bing+GPT it performs a little worse. An example that was simply better with Microsoft's Bing+GPT: "search GitHub for a Common Lisp project for performing SPARQL queries"<p>I periodically cancel my OpenAI ChatGPt Plus subscription because Microsoft's Bing+ChatGPT is good for research and generating code fragments.<p>EDIT: off topic, but I often prefer using GPT-4 via programs written in Python, Common Lisp, or Racket. Then, I can do web search and data/prompt preparation in my own code. For me the difference is: for research and coding help use Bing+GPT, otherwise write my own code.