3 Comments
Speaking for myself, I have yet to get correct information out of the playground, though I am confident that it will improve with time. I suspect it depends a lot on which functionality and toolboxes your question pertains to. I tend to ask a lot of questions about Deep Learning Toolbox functionality. Quite ironically, the playground's AI seems to know very little about Matlab's AI tools!
All that said, I would recommend cross-checking all code suggested by the AI by running it explicitly, and cross-check everything else with the documentation.
The answer accuracy will vary. For one, the AI model is trained with content up until June 2021. And your input prompt matters. Specific prompts with enough context will produce more accurate output. I recommend trying out and learning by experimenting. If you find any tips, please report back.