Strange that NVK was building some kind of an AI aggregator and recently talked about it on Livera’s podcast, but never bothered to run through Coldcard’s code with an LLM?

Even some of the earlier open-source Qwen models caught the entropy bug retroactively.