I have found ChatGPT to be more reliable than Claude, Gemni and other AIs in the heavy testing in the past week. It is not as pretty but it does the job more often than not.
This means that when you and another person are in a team it can be cheaper and more effective than an individual subscription. The Teams said it was more intelligent and so I tested it by using the same prompt as in my last article.
This time it gave me additional category of things to buy which was interesting and was longer. That is helpful for me because I like to experiment with how long I can make the output. How detailed can its answer be?
When I have gone back to test with Gemini I have been less impressed with it the more that I use it. It doesn’t seem to want to deal with files and output to many different file types. I am not being paid by MS. No one pays me for my opinions. I am just sharing what has worked for me. I will continue to test Claude/Gemni but now that I have the enhanced version of ChatGPT I hope it will be better. It is also possible that GPT5 will arrive this year or early next year. So that will keep things interesting.
I asked ChatGPT to make a picture appropriate of Alice in Wonderland and it acted like it was working and then said it couldn’t. When I asked why it couldn’t complete the task it said that it isn’t allowed to create things that would possibly infringe copyrights!! Can you believe that? It has no problem stealing from copyright holders, but it absolutely has the moral right and justification to say that stealing is wrong and it won’t do it. I might see if I can get it to recognize that itself is wrong for infringing copyright.
AI is a brittle intelligence that is just a guess at this point. Still it can save you some typing and for that it is worth exploring.