4 years ago we were asking DALL-E to make funny photos for us.

Now people can make an app with vibe coding in minutes.

I do think that we have overhyped what automation looks like in our lives, but it doesn’t mean it can’t serve a purpose. I just think we need to choose which tasks allow us to live life better.

First and foremost, if you can do things without AI and automate processes in your life you will actually save a lot more energy (both electrical and mental).

Yes your time is valuable, but if your idea of time well used is letting the golf course become the office while your stock portfolio works harder for you, we might have different conceptions of time well spent. While it isn’t bad to enjoy your life, with the estimated 80,000 hours in our working lives, you have something unique to contribute to society that brings out the human flourishing we’re all after. I believe automation as a concept can help refine your process of doing more of what you were made to do.

You will likely find an exhaustive list of online articles containing quick processes that will help refine your outlook calendar and minimize your inbox. I’m sure for many in corporate America, just a little more room to think would be nice. For me, as someone who’s been caught up in analysis paralysis way too frequently, I often think chat interfaces with AI give me more excuses to overthink something rather than taking action.

Thanks for reading! Subscribe for free to receive new posts and support my work.

So here are my top ways in which you can use automation to help you get to use your time in a more valuable fashion, coming from someone who chronically overthinks:

1. Write down your prompt before you ask AI.

This one comes from one of my favorite educational Youtubers, Network Chuck, when it comes down to organizing your thoughts. I have found this not only productive, but then I find out I don’t actually need AI to do the thing I’m thinking about!

2. Meta-prompt if you’re looking to offload mental deliberation.

If you absolutely need to prompt, I often find starting with that written-down prompt leads me to consider multiple angles I might need for an experiment. Well if you just ask the prompt to take on the role of multiple experts and have them “deliberate” in front of you, while you’re the fly on the wall, you can gain more perspective.

If ChatGPT is going to take up a lot of physical text,it might as well be worth reading.

3. Consider being wrong first and what it can teach you.

This goes back to number one. The current time we’re living in encourages us to use the fancy new tool first, but if you consider being wrong and have the courage to face the fear of failure (taking my own advice here), there is necessary growth in finding out what you didn’t know rather than relying on a safety blanket no matter what Claude might know.

So think of a problem → consider your context → do it as far as you can on your own → bring in AI once you’ve exhausted your physical capabilities.

4. Tell the AI to call you out.

I’m being dead serious. It will never be more helpful than telling you the absolute truth. Just be careful and don’t have it go off the deep end as Gizmodo tested in this article.

If you start with letting it “tell you where you’re wrong,” it will easily find areas for improvement in whatever you’re doing. But most of all, I have found that in recent years, the models have only grown to become more eager to please than to be 100% correct. So, with this behavior, have it skip the niceties and just tell you straight up what you’re missing at the end of every prompt.

If my agent resembles the likes of Bill Belichick in its tone instead of a therapist, I think we are on the road to continuous improvement.

Anything I missed? Feel free to comment below or share your favorite tips for automating the right things.

Enjoy the moment!