The conclusion here seems wrong:

> They found that ~70% copy-pasted the first output without editing it at all. .... This means that ChatGPT can immediately substitute for some workers. It’s a story of replacement rather than augmentation.

1. If they only accepted 70% and not 100% isn't that an argument for augmentation not automation?

2. Who wrote the prompt? We always need someone to write the prompt as these models are no generally intelligent agents.

I wrote a more nuanced take on this topic here: https://scalingknowledge.substack.com/p/why-job-displacement-predictions

Expand full comment
Apr 9Liked by Kenn So

Cool stuff, mate

Expand full comment