Your non-AI moat

"What's the Moat of your AI company?" That seems to be top of mind for founders pitching their novel idea to VCs -- and the most common question they get.

But I think it might not be the right question. Right now, nobody seems to have a good answer. a16z doesn't. At the same time, the pace of new builders piling into AI is dizzying. What if the answer is that AI tech is the wrong place to look for a moat?

Like, if you claim that Javascript makes your company defensible, you had better have an exotic and convincing argument for that. Most likely it doesn't; it's commodity tech. Similarly, the advantage of prompt engineering, or fine-tuning, or custom models, or datasets, as a moat is precarious: GPT-4 or 5 might just render all of this unnecessary, because users can just zero-shot whatever your whole application does, directly in ChatGPT.

Perhaps the right question is instead: knowing you're an AI company, where can you create something uniquely defensible, leveraging the fact that it is still early days? If you go through a list of the 7 Powers (Hamilton Helmer), then which of these do you have the most realistic chance of achieving, fast?

So: where's your non-AI related moat?