Email: Delete it All

Then use AI to retrieve what's important.

This idea is probably radical, but I’m building this solution anyway.

I’ve had it with inboxes and the flood of non-sensical, worthless, drivel that makes a mess of every workday. It’s a never-ending and consistently worsening scourge of digital life. Setting aside the long-standing quote I coined in the last decade of the previous century (“Email is Where Knowledge Goes to Die”), I need a clean break. I must end the madness.

So here’s what I did.

1. With a relatively clean inbox of about 300 recent messages representing current topics and conversations that I care about, I created an ontology¹ using generative AI².

2. Then, I deleted all those messages - inbox-zero.

3. Next, I eliminated the hundreds of Gmail filters I had accumulated over the years and replaced them all with one filter - when a message arrives - DELETE IT!

4. Finally, I enabled a custom Gmail service³ that monitors deleted messages every five minutes. It uses generative AI to classify each message within the ontology based on the message subject and body. When a message is detected that meets a similarity threshold, I move it into the inbox⁴.

The Grand Email Experiment

My approach assumes all email is spam. I want the email content to convince my AI process that it isn’t. And when it finds something noteworthy, it needs to decide if it should be (1) be pushed under my nose, (2) pushed into a digest, or (3) archived.

Perhaps I’m being too naive about this and underestimating the challenges while overestimating the capacity of generative AI to address this problem domain. But that’s part of learning the boundaries of AI.