A good next step is relevant, manageable, and clear enough that you will know whether you completed it. It should either produce something useful or answer a question that matters. Learning everything about AI or fixing your entire career is too broad to guide today's action.
A better step might be checking one training requirement, writing one honest experience example, or trying one task connected to an interesting role. Set a realistic time to review what you learned. Progress comes from connecting actions with evidence, then adjusting your direction as your understanding grows.