Browse Comments — Clean (de-noised)

This is the correct shift from model capabilities to structural execution, Rubén. A cloned product is a baseline, but the actual differentiator is managing the deployment reality and regulatory constraints. Complexity moving to the operational layer is where incumbents win.

I don’t think it can. Coding is not the differentiator

Rubén, AI replicates what you built, time replicates what you earned. Every moat on this list is a different kind of accumulated time. The founders who win in this cycle will be the ones who stop trying to outrun AI on features and start playing games where time is on their side.

Rubén Domínguez Ibar I am the Moat

Proprietary workflows and network effects tend to strengthen over time because they compound through continued usage and participation rather than through the codebase alone.

The "rebuilt in 48 hours" framing cuts both ways. Yes, a competitor can clone your surface in 48 hours. But I've shipped entire new products in 48 hours — because the reasoning infrastructure underneath is mine and took years to build. That's the actual moat: not the product, but the generative capacity that produces products faster than anyone can copy them. To answer your question directly: I'm building on a combination of #1 and an unlisted sixth — proprietary architecture protected by IP. Two granted US patents on compositional reasoning. A third filed last month covering reasoning over persistent heterogeneous knowledge graphs. Claude Code can rebuild a UI. It cannot rebuild the epistemological layer underneath it, because that layer doesn't exist anywhere else. Stack your moats, as you say. Mine are: hard-won architecture, granted patents, and a compounding data loop from every codebase Hokmah ingests. (hokmah.dev)

100% this! Everybody can copy you these days. Not everybody can make sure that their AI output is high quality though! That’s why investing in evaluation infrastructure is the smartest move to protect your business.

AI can definitely supplement proprietary data loops and make them self improving yet I agree on the other 4.

Spot on, Ruben. In a world where code is increasingly commoditized, your competitive advantage no longer lies in the 'what' (the product), but in your distribution, brand, and systemic execution. ​This is exactly why I prioritize acquiring established SaaS assets over building from scratch. Why exhaust yourself creating a product that can be copied, when you can acquire an asset that already owns a market, has a loyal customer base, and verified cash flow? Acquisition allows you to skip the risky validation phase and jump straight into optimization and scale. ​That’s the leverage I focus on via platforms like Flippa:

Google effectively will kill ai avatar companies and voice dictation companies with just one release , your moat points are valid but also fail against a behemoth like google who can use very tactic in the book to crush a product and will win too

Software as a medical device or Digital Therapeutics is an industry branch where this statement does not apply. An AI may copy the application, but it doesn’t have the clinical evidence which is needed for approval and reimbursement, so this particular market will not be at risk to get overrun by copycats

Honestly the moat I'd add is the 200 wrong versions you built before the right one. Claude Code can clone what you shipped. It can't clone the failed iterations that taught you why this version works and the others didn't. The code was never the expensive part - the decision history was.

The data loop one is the most underrated — most teams ship a thin wrapper around an LLM and call it a moat. Compounding usage > clever prompts, every time 👌

Very relevant. The easiest thing to copy today: UI + feature layer. The hardest thing to copy: Years of customer trust, usage patterns, internal workflows, and real-world distribution. A strong AI product isn’t just model + interface. It’s: AI + proprietary context + workflow integration + user trust + operational scale That’s where defensibility compounds. Great visual Rubén Domínguez Ibar

The moat people underestimate is institutional knowledge. You can RAG over a knowledge base but the actual edge cases, the "we tried that in 2019 and here's why it broke", the undocumented exceptions that senior folks just know, that stuff isn't written down anywhere. It lives in people's heads.

So these 5 elements can be traded and that’s where businesses would erupt, local LLMs trained which can forge partnerships

'governments take a decade to win' is a bit... non sequitur?

This is interesting, but not profound. The only thing that has changed is the scale at which it matters. Today, one person can match the pace of a traditional small team, but everyone has always had the ability to throw money at developers to have them create something. The cost of development has never been the only differentiating factor between competitors.

the data loop one hits different when you're building something with real user-contributed data. that moat grows every single day and nobody can just copy it overnight

Our competitors cannot rebuild our product within 48 hours with Claude Code, because Claude Code will not manage their servers and they will run out of money in 1 day.