When I'm in ChatGPT (my personal assistant), I am stern. Authoritative. If not, It will only tell you what you want to hear. Like the CEO, he does not have time for sugar coated answers and feel good solutions. I.E. Your task is too .......Take on the role of a.......I expect these results.....NO, thats wrong, re-evaluate and perform a deeper analysis that will bring results that will meet my expectations and follow my instructions very carefully. If you have any questions, ask, do not assume anything.
When I'm in CLAUDE (my technical expert) we work symbiotically. I.E. Have a look at this Code. Identify any errors and inconsistencies and report back to me with potential solutions for my review. Sternness not necessary, it's mutual respect. An intelligent conversation with a trusted expert.
With GROK (my research assistant) I'm like, ummm, would it be ok if, like, you went through the steps of installing python on my new laptop? I am kind of slow so if you would give me step by step instructions click by click, I would greatly appreciate it. And if it's OK with you, would you please stop after every step and not proceed to the next step until I say next, would that be OK? I am so sorry, I am just not as fast as I used to be at all this stuff. You are such an awesome teacher.
I AM TELLING YOU! GROK will shut you down in a heartbeat if you raise your voice. Certain words that show frustration and any disrespect will shut down the chat in a heartbeat. I am positive that Elon Musk has specifically told the dev team to program certain rules into the algorithm. He does it on purpose. Especially on free accounts.
perplexity? stay away from it. It is spyware (IMHO) and has infected every device I have ever tried it on. As soon as I deleted it, everything went back to normal. (Think about it, why do they give its $20 a month service away for free for a year...on a regular basis? I know why, it aint good). Grok blows away perplexity anyway, you just have to play nice. LOL