You know that kid who annoyingly crushed every test? Well, ChatGPT creator OpenAI just unveiled its most advanced AI model yet - GPT4 - and it puts that kid to shame.
...with one exception: English tests.
According to a new whitepaper, the model scored in the 90th percentile for the LSAT, the Bar, and various SAT components. It even placed in the 99th percentile for the Biology Olympiad. But despite its prowess, on AP English Lit and AP English Language tests, it landed in the 8th and 14th percentile.
In other words, the cliché of the writing and reading comprehension being the least employable skillsets appears to be aging poorly.
Will subsequent artificial intelligence models eventually master the infinite nuances of the written word?
Maybe. Probably.
But also maybe not. As author W. Somerset Maugham put it, "There are three rules for writing the novel - unfortunately, no one knows what they are."
Without rules, AI is a sad robot!
|