Pretty neat. I bet were going to see a lot of GPT-3 powered school work in the future. "Write me a turn paper on..." and then your just cleaning it up and making adjustments.
I bet a lot of people would get away with it, too. These generated texts are often just convincing enough to make you think they would make sense if you were less tired/distracted/whatever. A teacher in a rush or at the end of the day could easily be fooled (especially considering the quality of many human-generated term papers).
Now we need an equivalent of a "deepfake recognizer" for these sort of generated texts, to distinguish texts with some sort of coherent meaning from texts without.
As a bonus, such a model could be used in a word-processor to redline places where a human writer has written a lot without saying much.
This presentation almost definitely passes the Turing test: if it were posted here without any further context, people would definitely be arguing about whether coding was really "so simple even a monkey could do it".
It started by saying how to be a coder in Japan, then doesn't mention Japan ever again. All coding problems are exactly the same until a few slides later when actually they are complex. The paraphrase of the feynman quote isn't actually a paraphrase at all. There's no actual substance. It's like a train of thought which gets derailed every 5 seconds.
When I was in school in the 00's I had a lot of classes around how to write search queries.
I feel like the way this GPT-3 query was phrased was rather clever, and I wouldn't be surprised if we saw "how to prompt GPT-3" as the assignment itself in the medium-term future.
