Surprisingly Good AI-written Content

Recently, I've happened upon a few pieces of AI-written content that I have been surprised and delighted by. There's no doubt that the content was written by AI, but when I read it, it didn't feel AI-generated. In the land of AI slop, it was refreshing.

Some observations:

  1. In both examples, the website creators make clear indications that AI is used to generate articles

  2. Both leveraged AI's ability to do deep research, gather information from a wide variety of sources, and cross-reference across source material to form a unified analysis

  3. Both were extremely accurate portrayals / descriptions of myself and my company

  4. I was left not only wanting more content, but, in classic engineering fashion, scratching my head about the prompts used.

Example 1: Lenny Zeltser's RSAC Innovation Sandbox profiles

https://zeltser.com/media/rsac-2026-sandbox & https://zeltser.com/media/rsac-2026-sandbox/clearly-ai

Lenny wrote an incredible analysis of the top 10 finalists for RSAC Innovation Sandbox 2026. Or, at least, his MCP server did.

The analysis was performed by AI without direct human validation, to demonstrate the capabilities of AI agents guided by an expert framework. Outside this demo, a human analyst would conduct conversations with the AI agent to arrive at more accurate conclusions.

And it was impressive. I deeply enjoyed the analysis of Clearly AI, the knowledge about our company profile, and how we compared to the other finalists.

Example 2: "YesPress" Founder Profile

https://yespress.io/emily-choi-greene

I'm not 100% sure what this "newsroom" is, or what their goals are, even after perusing on their site for 30 mins after reading my own feature.

It was written with a human voice, with interesting infographics, and a scarily old photo of me (why that photo?!) If it wasn't published in 2026 alongside some 100 other features, I could've been convinced that a human wrote it. (Depressingly, all content on the internet published after ~2024 I have trouble trusting is human-written anymore. Maybe more depressing, it clearly read this whole blog, which was written by a human for humans but in reality is probably only read by AI.)

This article was uncanny, especially because it wasn't clear why *I* was selected by their AI as a feature (or did the human pick the subjects?) I'm only one of a few YC S24 founders to be featured, for example.

Also, their X profile (which tagged me, and is how I found out about this at all), is followed by Obama.

For occasional essays on building, sign up

For occasional essays on building, sign up