Treason and Robots

[sdi]

I started studying William Linney's Getting Started With Ancient Greek the other day. There's a famous Italian adage, "traduttore, traditore"—in English, we might say "translation is treason"—and the biggest hurdle I've had in studying the classics is that I've had to contend with the biases of translators. Most of them clearly don't understand the material, but even Thomas Taylor—bless him—is so obsessed with Proclus that he reads Proclus, and writes Proclus, into everyone. So being able to bypass such gatekeeping is valuable. But to be honest, I'd content myself with translation if there were translations of everything I wanted to read—but I really want to read Stobæus, and nobody's translated him!

But that's not what this post is about: this post is about blind faith. I was mentioning the above to a friend, who asked, in all seriousness, "Why not just ask ChatGPT to translate him for you?"

I stared, incredulously, for a moment, and answered, "How would I know if it was right?"

He was quite surprised by this and conceded the point.

Now, this seems a very obvious thing to me—the whole point was that I can't trust a translator of any kind, human or robot—but it's not as if my friend is a dummy! Rather, it seems there's something of an insidious meme of the infallibility or inevitability of machine learning which is polluting people's abilities to think clearly about it. I think it's worth bearing in mind that a meme is all it is, and this meme is pushed by people who are neither honest nor have your best interests at heart.

Remember that your highest self is essentially independent. To be your best, think for yourself!