Good Ideas are Your Only Moat Now

In an age of AI-driven commoditization, the most valuable work is the deep thinking that happens before you prompt

I’ve been thinking a lot about where good ideas come from and why they seem to matter now more than ever. The pace of AI progress is accelerating at a dizzying rate, and it’s creating a new reality for builders. For years, a company’s moat might have been its proprietary code rooted in a novel way to accomplish something, or the sheer time and effort it took to build a complex feature. But that’s changing.

AI coding means development can happen faster and faster and capabilities that were once a company’s “secret sauce” can now become a single API call. In fact, this second scenario has become so common that it’s now one of the core questions I ask PM candidates: “What do you do when a new model update commoditizes your magic?”

This shift means the value is moving away from the technical act of building (which is now tablestakes), and toward the quality of the idea being built. In a world of content abundance and near-zero cost of creation, your ability to generate novel, insightful, and well-reasoned ideas isn’t just a business asset; it’s your most valuable marketable skill.

This new reality was recently highlighted in the AI product community when OpenAI launched ChatGPT Pulse. The team at Huxe, a startup that had been working on a similar concept, saw elements of their work echoed in a major launch. As founder Raiza Martin shared, it was a powerful reminder of the constant, demanding need to generate new ideas.

But it’s important to note that the core concept of a proactive “daily digest” isn’t entirely new. Many of us remember Google Now, which, back in 2012, used our data to surface predictive “cards” with things like flight information, traffic updates for our commute, and real-time sports scores. The idea was powerful then, and it’s powerful now.

The difference today is the execution. With modern AI, the potential for personalization, synthesis, and the sheer quality of the insights is an order of magnitude greater. This highlights a critical point: when execution becomes easier, your durable advantage comes not just from the novelty of your idea, but from the quality of your thinking and the thoughtfulness of the experience you build on top of it.

With the pace things are accelerating these days, it also means that a single brilliant idea, even when perfectly executed, is no longer enough. The real, lasting advantage comes from the ability to consistently generate exciting new ideas, staying one step ahead of the commoditization curve.

This raises a more profound question. As AI gets better at executing, are we at risk of outsourcing the very thing that matters most - our ability to think deeply?

The Threat of Outsourcing Our Brains

I was captivated by a recent, thoughtful essay from Derek Thompson, “You have 18 months,” where he makes a provocative argument. The real threat of AI, he posits, isn’t that it will take our jobs, but that it will degrade our own capabilities by making it too easy for us to stop using our own minds.

Several of his points rang especially true for me, crystallizing ideas I’ve been wrestling with both as a PM and a parent.

First, he champions the importance of long-form reading. This is something I’m actively trying to do more of, and it’s a skill I want to ensure my kids cultivate as they get older. In a world of infinite, bite-sized content, the ability to sit with a complex text feels more important than ever.

He connects this directly to another idea I loved: reading as the ideal practice for systems-level thinking. Systems Thinking is a quality I believe is absolutely essential for great product builders - it’s even one of the core characteristics I look for when hiring an AI Product Manager. The ability to hold conflicting ideas in your head, to see how different parts of a complex system interact, then connect the dots to form new ideas and novel concepts is a muscle, and deep reading is a key tool in helping you train it.

Finally, the essay perfectly captured the art of connecting dots across different domains. I often get asked “Where did you read that?” or “How did you come up with that idea?” The answer is rarely a single source. More often, an idea is the result of braiding together threads from a Substack post I read, a deep-dive podcast I listened to on my commute, a conversation with a founder, and a random thought from one of my own side projects. It’s this synthesis - this “time under tension,” as Thompson calls it - that generates something combinatorially new.

If we offload all of our writing to AI and replace deep reading with algorithmically-fed summaries, we risk losing the very practice that enables this kind of synthesis. We risk unwiring our cognitive superpowers at the very moment we need them most.

My Playbook for Cultivating Ideas

This has made me reflect deeper on my own process. Where do my ideas come from? I’ve realized it’s not a single source, but a system of inputs and practices designed to keep me in a state of constant learning and connection. It’s my personal method for maintaining “time under tension.”

1. Consume Widely and Deeply

• Deep Dives: I make it a point to read thoughtful Substack posts and articles regularly, and as Derek Thompson’s piece reminded me, I’m always trying to prioritize more long-form books.

• Active Listening: My husband and I listen to audiobooks and deep-dive podcasts like Acquired during our commute. My favorite part is pausing to debate and discuss what we’ve just heard, which helps solidify the ideas.

• Real-Time Pulse: I also stay on top of emerging topics and real-time trends by reading and engaging with content on platforms like X, Reddit, and LinkedIn. This balance between deep focus and the fast-moving pulse of the industry is crucial.

2. Connect Through Conversation

• Structured & Unstructured Chats: I set aside time on Fridays for 1:1s with people inside and outside my team and even field. Just as important are the impromptu “water cooler” chats in our micro-kitchens at Google, where some of the best, most unexpected ideas are sparked.

• Industry Events: I try to go to at least one external event a month, from large conferences to small, invite-only dinners. The energy and perspective from other builders is invaluable.

3. Create Through Play & Practice

• Side Projects: My 10+ AI side projects are my personal lab. They are an excuse to play, to get my hands dirty, and to bump into the limitations and possibilities of the tools myself.

• Writing to Think: This newsletter is my forcing function. The act of writing forces me to clarify my own thinking and connect disparate dots into a cohesive narrative.

• Presenting to Solidify: Giving talks forces me to structure a story and defend it, which is the ultimate test of a well-formed idea.

In an age where AI can provide answers instantly, the most valuable human work is the slow, deliberate, and sometimes uncomfortable process of asking the right questions and actually taking the time to try answering them. The future will be built by those who don’t just know how to prompt, but who invest in the deep thinking that makes a prompt worth writing in the first place.