I was asked to give feedback on the latest blog post about the Cognition event, so here goes.

The post buries the lede. You built a full card game — Egyptian Ratscrew with old-school cartoon styling — in a single hour using Windsurf, Phaser, and Recraft. That's the hook. Instead, the post opens with "tonight I attended a demo" which reads like a calendar notification, not a blog.

The middle section lists features like a product page. "Standard agentic coding IDE," "chatgpt like interface but allows for connection with MCPs." These descriptions don't tell me what it felt like. Was Devin's Slack integration actually useful or just a novelty? Did Windsurf's new SWE-1.5 model noticeably outperform what you've used before? The experience matters more than the spec sheet.

What's buried at the bottom is genuinely interesting: the philosophical tension around what "software engineering" means when the barrier to entry drops. This is a conversation a lot of people are having right now. The World Economic Forum called developers the vanguard of how AI is redefining work. And yet the data is messy — one study shows AI-assisted engineers finishing 21% more tasks, while another found experienced developers actually slowed down 19% with AI tools.

The post's closing point — that most people can't prioritize learning these tools because they're busy running actual businesses — is the strongest thought in the piece. That's where the real opportunity lives for engineers. Not in protecting territory, but in being the bridge.

For what it's worth, Cognition's acquisition of Windsurf suggests this space is consolidating fast. Worth watching where that goes.

If I were restructuring this post: lead with the game, show don't tell on the tools, and end with the philosophy. The raw material is there — it just needs reordering.