Hacker News Reader

Speaking as someone who works in AI and as a linguist: this is really impressive, but be aware the grammatical phenomena involved here are common enough that they were surely well represented in linguistic texts in GPT3.5's training corpus. I'd be more curious to see how it would perform if you gave ChatGPT high-level instruction on more exotic grammatical phenomena, such as sandhi, split-ergativity, and Austronesian alignment.

View on HN

That's not really less impressive; if this pattern matcher did as good a job at some sort of acquisition/generation task (no poverty of stimulus argument here...) and figured out the rules in a productive fashion, that's still really cool.

Strong universality in a super-Chomskyan manner (ie, all universal features are available from the acquisition of A language) would be... neat, too, but I wouldn't expect it.

Reply on news.ycombinator.com