That made me laugh a bit as well. Definitely want to see some rigorous testing on that, I'd expect that on longer calls tha caller can make the ai say basically anything.
That requires a GitHub account, which most readers on most blogs won't have. Moreover, I'm pretty sure most people (>50%) who would have commented, probably won't do it if it requires creating an account for an external service first.
Maybe anonymous commenting by just solving some captcha is impossible nowadays because in practice all captchas get broken by spam bots immediately. Or maybe not. I would like to see evidence for that first before giving up and choosing a commenting solution which strictly requires an account.
