95: How you can use agentic AI to earn more money and win back time

Summary: In this episode Dan explores agentic AI, how it differs from chat-based AI, and why delegating to it rather than using it can help you earn more and reclaim time. 5 things you'll learn in this episode The key difference between chat-based and agentic AI. Most people are directing AI step by step. Agentic AI works differently: you give it an outcome and it figures out the process, much like delegating to a trusted person rather than micromanaging every move. A practical filter for deciding what to hand over. Any recurring task, anything involving multiple tabs or tools, and anything that doesn't require knowledge only you possess are now prime candidates to delegate. This alone could reshape how you spend your days. What to keep for yourself, and why it matters. Writing your own articles, generating ideas, doing creative work. The benefit comes from the doing, not just the output. Knowing where to draw that line is just as important as knowing what to delegate. 13+ real examples of tasks being delegated right now. From content suites and podcast show notes to Canva assets, email campaigns, and data organisation. These are live examples from the past few weeks. Where the real opportunity lies. It's not just about saving time. It's about what you do with the time you get back. When the tasks that don't require you are removed, what remains is the work that genuinely moves things forward. How most people are currently using AI At the moment, most people are using AI in what I'd call a chat-based way. You open up something like ChatGPT, Claude, or Gemini, you type in a prompt, and it gives you a response. Then you type another prompt, and it gives you another response. And so on. It's essentially a series of individual tasks. You are directing every step of the process. You are telling it what to do, one instruction at a time, and it is responding to each of those instructions. There's nothing wrong with this. It's incredibly useful, and for a lot of people it's already a big step forward from how they were working previously. But it does have a limitation. You are still doing the thinking around how to get to the outcome. You are still breaking the task down into steps. You are still effectively managing the process. The shift to agentic AI Agentic AI works differently. Instead of giving it a series of steps, you give it an outcome. You don't say, 'do this, then do this, then do this'. You say, 'this is what I want done', and it works out the steps required to get there. The way I've found it easiest to think about is to compare it to working with another person. If you were delegating a task to someone and you had to micromanage every step, telling them exactly what to do at each stage, that would feel very similar to using chat-based AI. But if you were working with someone you trusted, you would simply give them the job and the desired outcome, and allow them to figure out the process themselves. That's much closer to what agentic AI is doing. It's not just responding to prompts. It's completing tasks. What this actually means in practice Once you start thinking about it this way, the question becomes less about 'what can AI help me with?' and more about 'what should I no longer be doing myself?'. For me, a useful filter has been this: Any repetitive or recurring task, any task that involves moving between multiple tabs or pieces of software, any task that doesn't require knowledge that only exists in my brain, and generally anything that would take me somewhere between 30 minutes and a couple of hours, those are now candidates to be delegated. What remains is the work that actually requires me. Thinking, decision-making, coming up with ideas, and doing the parts of the job that are inherently human. A quick note on what I'm not outsourcing This is important, because I think there's a temptation to go too far with this. I still write my own articles. I still come up with my own ideas. I