I’m still looking for no-moatists to logically explain why there is no moat in SOTA LLMS.
That paper released by a lone Google employee 3 years ago still misguiding people it seems.
My reasoning is that we have seen multiple small labs come out of nowhere and create impressive models.
The vast majority of people simply don't need SOTA models. I run most of my programming on a single 7900 XTX with Qwen 3.8 27b. I'm more interested in scaling out to run more instances of the model in parallel, than I am in scaling up to better models.