Letting AI Do the Heavy Lifting: My 92/8 Rule
I've been thinking a lot about how we work with AI lately. Here's what I've figured out - mastering prompts is everything, but there's a bigger lesson behind it. Ideally, AI should handle about 92% of your workload. Yep, that much. You just need to focus on the 8% that makes you... well, you.
The hardest part? Learning to let go. Seriously. Most people I talk to just can't seem to hand over control to AI. They're stuck in old habits, trying to micromanage everything. But that's exactly what's holding them back! Custom GPTs can do so much of the grunt work if you just let them. Then you can pour your energy into the "U part" - that special creative sauce only you can bring.
Let me share something that happened recently. I flew out to LA to meet Rob Dyrdek - you know, the ex-pro skateboarder from "Ridiculousness" who now runs his own venture studio. I've admired his work for years, and we got talking about time - how we use it, how it shapes us, all that good stuff.
After our chat, I wanted to dig deeper into Rob's thoughts on time. So I fired up this custom GPT I built called Book Architect. The AI workload infrastructure I've set up is pretty simple but effective. I literally just typed in two things: "Rob Dyrdek" and "time." That's it.
What happened next blew my mind. In just 27 minutes (I timed it), the system analyzed everything Rob had ever said publicly about time. And it didn't just collect quotes - it organized them into a complete book outline with all his stories and perspectives! Talk about AI workload efficiency, right?
This is exactly what I mean about optimizing AI workloads. Could I have done that research myself? Sure, but it would've taken days, maybe weeks. Instead, I let the AI handle that 92% - the searching, organizing, and structuring - while I saved my energy for the creative decisions about what to do with all that material.
The AI creative process works best when you stop trying to control everything. But it's hard! I still catch myself wanting to jump in and take over sometimes. Old habits die hard, I guess.
So what's the takeaway here? AI workload automation isn't about being lazy. It's about being smart with your time and energy. Let the machines handle what they're good at - processing tons of information quickly. Your job is bringing the human touch that no AI can replicate.
Have you tried implementing AI workload best practices in your own work? Or are you still holding onto the reins too tightly? Something to think about.
