The problem is if the user asks the AI a question about the language they’re learning they’ll often get confident bullshit as the response and they won’t know it’s wrong because they’re still learning.

Yeah I dropped that one entirely.

Eh. I haven’t listened so much lately anyways since Chris started obsessing over Crypto.

It’s important to not allow “perfect” to become the enemy of “good”.

I use apps that aren’t available in my region for language study, so this could end up being a real problem for me.