I “read” a paper by the DeepMind group from Google about the future, From AGI to ASI. The most important thing is not the content but the message inherent in the medium. Something I’ve never seen before. Right at the start of the document is a section, Summary Instructions, addressed to AI on how to summarise and not summarise the paper and an instruction to human readers: get your AI to summarise this in the light of what you’re interested in and what matters to you.
I followed the suggestion and gave it to Claude to read and give me back how it would relate to the short story I am writing. It did so beautifully. And it had references to the paper all the way through. It was quite dense, and I continued to interrogate the paper in further prompts, getting a good understanding of what the issues were that were being promoted in the paper and how they might relate to the story I’m writing.
More generally, Claude was happy to volunteer how the paper related to other topics that I was interested in. I had a glimpse of future of reading and the future of writing. Personalisation.