Building Nyantrace.ai · NYU Tandon
I'm building Nyantrace.ai — observability for AI agents. These entries are where I think through AI, platform economics, and what it means to build right now.
Poe Poe Myint Swe →"94% theoretical vs 33% actual. The gap between could and does is where every adoption framework, every change management playbook, every Davenport-style quick win lives..."
"The adoption playbook is universal. Center of excellence → quick win → scale. The technology changes; the pattern doesn't. But ecosystem adoption is fundamentally harder..."
"The product IS data — and the agent IS the interface. UI goes from being a product to a byproduct. In the extreme future, the agent becomes the entire experience layer..."
"AI platforms are the new gatekeepers. Distribution is the new moat. Whether your brand is old or new doesn't matter — what matters is whether the AI picks you..."
"Disruption has a specific meaning — and Uber doesn't qualify. But the theory is incomplete: innovation also means addressing opportunity and need gaps..."
"Don't start with moon shots. Of 152 AI projects studied, the ambitious ones failed while the low-hanging fruit transformed businesses..."
"There's no value in the stuff your company offers. Value lives in what your customers can DO with the stuff you offer..."
"There's no reason why you and I have to have the same store in the digital world. Technology makes it easy to build a million different stores..."
"Strategy is not about what you do in 2030—it's about the projects you implement today across three boxes..."
"AI-assisted development removes the engineering gatekeeper—not by replacing engineers, but by amplifying what one engineer can build..."
"We face algorithmic feudalism where AI perfects rent extraction, markets die, and platform lords become permanent..."