It's a good starting point. It may get some things wrong, but that's where my ability to research and truth seeking lies. I usually google more anyway because i want to go deeper than what chatgpt offers. I will eventually get to the truth. But the basic questions are well answered. I once asked it to explain the whole "Field has momentum" theory in a way a Feynman lecture would. It did it surprisingly well. I had some background so I kept asking more questions, and it was alright. Augments the curiosity, and make the topic seem simpler. Tools I use are: Google, science publications, and archive to see if there is text around it.
One small thing, the idea that chatgpt will feed bullshit is possible, so you have to be a bit skeptic and check the info, but it gets things right most of the times. Fun fact: I confused chatgpt about static electricity ie "Why does my hair get a static charge by rubbing a balloon?" It kept saying positive charge or negative charge, with transfer of electrons. Truth is, it's still poorly understood. [1] this paper highlights the conundrum.
For a beginner in college level physics, chatgpt is an excellent resource for understanding concepts.
[1]: https://www.researchgate.net/publication/360674587_Derivatio...