When we say a model is intelligent. What does intelligence mean? What does reasoning means? What is intelligent questions? The right questioning starts from the very fundamentals, the What ,which ai model does this better,which ai model sticks to the question to the very end to figure out the truth or discover sth critically new? Does models have curiosity? What is curiosity? Does intelligence only means that the model was able to perform task it was told to do by figuring it out? Is that all there to intelligence? Is benchmarking really a way to measure it? How benchmarks are created and tested on in the first place?.

If you find a piece of paragraph of a particular subject and paste it to the ai, and prompt it to ask question , how does it perform? Do models ask questions? How much questions do they ask? Does the quality of those questions measured? Do we specifically train ai models to ask questions? How can we do that?how can we find quality sets of questions? And is the dataset the only way to do that?.

Can model ever develop curiosity?why we ask questions? Why would ai ask questions? Asking right kind of questions opens up the brain a lot and drives thinking to right place, does that mean if an i model can actually ask right kind of questions will it become intelligent on its own?will it help the ai model to think in the right direction? How reasoning inside ai models happens in the first place? Do you guys ever see what kind of questions the ai model asks ,did you guys ever evaluate them? Or experiment simply to see what kind of question will it ask on a particular subject for discovering sth? are you guys curious to test these things? Anybody else have done these things? Or curious to talk about it ?or have noticed or observed sth valuable from your own experience to share here