Key Takeaways

  • Alex Cornell connected an AI agent named Marlo to his Instagram account to sort thousands of forgotten bookmarks into custom categories, a personal golf manual, and a retro SkyMall catalog.
  • Instead of letting old writing sit idle, Cornell scraped 20 years of his Blogger and Tumblr archives, turned them into a central document, and set up an agent to read random entries back to him as a daily morning audio show.
  • Claire Vo noted that consumer agents with visual web connectors unlock creative workflows because they reason directly over visual interfaces rather than structured API endpoints.
  • Effective consumer agents mirror messaging apps instead of developer consoles, using progressive disclosure to hide background operations until a user asks to see them.

The Graveyard of Personal Bookmarks

Everyone hoards digital content they never revisit. You bookmark technical threads on social platforms, save training tutorials, and store recipes in native app folders. Those bookmarks disappear into an endless scroll. The friction of organizing thousands of posts manually ensures they stay buried forever.

Cornell attacked this issue by pointing an agent at his personal accounts. As he explained: “I had Marlo look at everything I'd saved on Instagram and just organize them. So, put all of my saves into folders and categories.” Once the agent held the organized index, Cornell asked it to synthesize the contents based on his actual interests. It extracted every golf tip he had ever saved and compiled them into an instructional handbook. “I had Marlo do it and then consolidate them all into like a manual and so that I had this like kind of like for Alex like these are the things you should work apparently you think you should work on,” Cornell recalled.

Turning Archives into Living Formats

Raw archives fail because reading through old databases feels like homework. The solution is formatting the retrieved data into an entertaining medium that fits existing morning habits. Cornell took his 20-year writing history across Blogger and Tumblr and fed the complete collection into his agent pipeline.

“So it went back and found all the posts and then accumulated them into one PDF first, then this little thing that is basically just like looks like a blog. But the way that I like to read it now, just again for memory, is via a podcast, which is sent to me every morning, and it is just basically a random reading,” Cornell said.

He applied the same design instinct to commercial saves, converting saved product posts into a 1990s-style SkyMall shopping catalog. As Cornell put it: “one of my favorite things, I think, to do with these systems is take what would be otherwise like a fairly mundane set of data and render it in a way that I personally find really fun.”

Vo highlighted why this works technically: “I think the ability for these agents to reason with visual data in a way especially if you can just log into a website or use an Instagram connector just really unleashes a lot of creativity use cases I feel like have been harder to do before.” When agents can see and click through existing consumer interfaces, users do not need clean APIs to extract their trapped data.

What to Do With This

Pick one platform where you have accumulated hundreds of saves, such as your browser bookmarks or saved social posts. Export that archive to a single text or visual folder, connect it to a multimodal agent, and prompt the agent to produce a single two-page reference document on your most common topic this weekend.