I am curious how other programmers that I have no interaction with in this subject , are using LLM AI tools in their daily programming workflow.
Although I use them I have an issue recommending them to others in my company because I believe you have to have a certain amount of experience to understand the B..S.. parts from the real useful ones. The coding hallucinations are starting to become so great that you can't tell if a library that the code uses doesn't exist (and never existed) because the usage and references are plausible. When you ask general programming questions you usually get answers that are common misconceptions and far from the truth. e.g. Some moments ago Gemini told me that when it replied that Memory tables are not faster than MyIssam or InnoDb it meant that most servers have less RAM than the table size !!! (maybe that was accurate 15 years ago ... but again ... not even then ... )
But... I realize that more and more ChatGPT and maybe even more Gemini , are part of my starting point when I search something related to programming.
What is your experience so far ?