As new AI tools start to introduce monthly fees, with ChatGPT being over $40, a divide between those who can afford to pay and take advantage of automating and augmenting aspects of their work and those who can't is just going to push an even wider wedge into the gap.
I've seen a lot of people use GitHub's copilot to write tests for a given function. I don't think this is necessarily a bad thing; writing test can be laborious but also helpful for discovering edge cases, but the developer who is paying to generate them is going to appear more knowledgeable and productive than someone who needs to keep writing them manually.
I also don't think we're in a place where AI generated code can be assumed to be good or better. The developer still needs to be able to weed out the bad ideas. For now.