Quick answer
An internal Google document, leaked in 2023, became known as the 'no moat' memo. It argued that proprietary AI companies, like Google and OpenAI, lacked a sustainable competitive edge. The memo claimed open-source AI models, which were rapidly improving and becoming cheaper to run, posed the true threat. Open solutions, it suggested, were quickly catching up to closed ones, blurring the lines of advantage for industry giants.
The Memo That Shook the Status Quo
Imagine an internal memo from a major tech company. It wasn't worried about its direct competitors, but about something far more foundational. That's exactly what happened in February 2023.
A Google document, reportedly from one of its researchers, found its way into public view. Its central claim was stark: "We have no moat, and neither does OpenAI."
This wasn't just casual chatter. It was a serious assessment of the AI landscape. In business, a "moat" refers to a sustainable competitive advantage. This protection guards a company's profits and market share. Think of strong brand loyalty, unique patents, or a massive network effect.
The memo suggested this protective barrier was rapidly eroding, especially for large language models and generative AI. The real challenge, it argued, came not from another corporate giant, but from the open-source community.
Open Source: The Unseen Challenger Emerges
The "no moat" argument's core rested on the rapid advancements within open-source AI. Historically, developing sophisticated AI models required immense computing power, vast datasets, and highly specialised talent. Most of this was concentrated within a few large corporations. This naturally created a barrier to entry. But things were changing fast.