General advice - if you don't have a very specific reason for doing a PhD e.g. to do a post-doc or stay in academia or really want to dig deep into a topic due to personal reasons/drive - I would rather not start a PhD.

If you don't want a 'meh' PhD you must either be really good and find a hands off supervisor or find a really good supervisor that is willing to work with you.

You can work in the industry on ML topics with just a master's, you just have to show that you can get things done.