Of Moats and Ramparts: Rethinking the innovator’s dilemma in the age of AI
June 2025
Alline Akintore
In Clayton Christensen’s 1997 book, The Innovator’s Dilemma, the author gave the startup world a framework that shaped generations of founders: disrupt the incumbents by starting small, serving overlooked and underserved users, and climbing steadily upmarket, armed with technical differentiation.
Today the paradigm is shifting. For generative AI applications, moats are evaporating faster than you can say “fine-tuned LLM.” What used to take years of R&D and infrastructure can now be prototyped over a weekend with GPT-4, a Mistral model, and a well-thought-out prompt. Founders are shipping MVPs at warp speed—and they’re not the only ones.
Incumbents, once slow to react, are not sitting out this platform shift. Large players are deploying capital, teams, and compute into AI, with access to the same models as the startups trying to dethrone them. Yes, bureaucracy and brand risk still slow them down—but they’re in the arena now, and the early-mover advantage is thinner than ever.
Which brings up the question du jour: what does the innovator’s dilemma look like when everyone – both incumbents and upstarts – is innovating on a level playing field with the same tools? Here is our take on where founders should be focusing to build durable companies at a time when the tech advantage is thin and fleeting:
Use case depth: It should go without saying, but solve a real pain point for your customer. In a market flooded with upstarts built on the same models, the winners will be the ones who go deep and not just wide. This is the time to care obsessively about workflows, edge cases, and actual pain.
Own the workflow: Build solutions that people rely on. The tighter your product is stitched into your users’ habits and behaviors, the harder it will be to be ripped out. With the deluge of AI tools and solutions, this will be key to defensibility.
Lock in distribution, distribution, distribution and optimize speed to market: whether through partnerships, integrations or community. In this world where differentiation fades fast, getting to market first and learning from users is key.
Proprietary data loops: the best moat might not be your code or UI but it could be your data. Feedback loops from product usage and access to proprietary data, particularly in vertically integrated apps, will separate your solution from the pack. Lean into this.
This version of the innovator’s dilemma asks not “how do I catch up to the incumbents?” but “how do I stick when everyone has access to the same magic?”
The good news: if you're a founder in Oregon or the Pacific Northwest, you’re building in a region that understands both grit and reinvention. If we witnessed anything during Portland’s recent Startup Week, it is that this is a place where new ideas are meant to take root, and we are riding this AI platform shift full throttle.
If you're a founder here in Oregon or Southern Washington and you’d like to add to (or challenge this view), we’d love to hear from you.