The Myth of the AI "Moat"
And Where Real Value Lies
The tech world’s obsession with finding an AI "moat" has reached fever pitch. First, it was OpenAI’s perceived advantage with proprietary models. Then Meta’s Llama models disrupted the narrative by open-sourcing powerful alternatives. Next came debates about hardware supremacy – like DeepSeek’s breakthroughs in optimizing NVIDIA GPUs by bypassing CUDA layers. Now discussions revolve around inference efficiency and cost optimization.
Why are we so obsessed with the idea of a moat?
It seems to stem from this race in AI where everyone wants to be the leader. Personally, I think that approach is misguided. I don't believe there’s an actual moat in AI if we're talking about a winner-takes-all strategy. In the future, AI will be infused into almost every product and service, the same way everything became digitized. And historically, tech costs always go down over time.
However, there is a moat in terms of a key differentiator—particularly the unique advantage you can build within your own business.
My Time at SAP’s Business Unleashed
Last week, I spent time with the team at SAP during their Business Unleashed event, where they unveiled their new Business Data Cloud in partnership with Databricks. One thing that really stood out was how clearly they articulated their own “moat” and, more importantly, how they plan to help other businesses establish theirs.
It all comes down to the data that companies have—especially the semantic layer of data stored in SAP systems. I spent two days listening to the executive team discuss their products and services. But it only took about an hour of hearing Mohammed Alam, head of the SAP Product & Engineering Board, to realize the big picture: your real value is in your data, and a lot of that data is trapped in multiple systems. For many enterprises, one of those major systems is SAP, which often holds critical information like finance, supply chain, or HR data.
Get Back to Basics
What I love about SAP’s approach is that it’s not just flashy marketing. They’re focused on the core of where the value lies: the data. So, if you want to be a leader, don’t get tangled up in trying to build your own moat from scratch. Instead, get back to the essentials:
Get your data in order.
Modernize your systems in the cloud so everything connects seamlessly.
Speed up your prototyping with clear, core use cases that leverage your unique data.
Lean on the platforms and providers that have done the heavy lifting—like SAP.
In other words, stop trying to build an elusive moat and focus on the fundamentals. If you’re interested in learning more, I had the chance to sit down with Irfan Khan, the Chief Product Officer at SAP, to talk about the release of the Business Data Cloud. I encourage you to check out that interview for deeper insights.
