How to read (and actually retain) non-fiction using AI tools

Last year, I read 25 books, 14 of which were non-fiction. I was happy about the number but the frustrating part was retention -- a few weeks after finishing a brilliant book, the core ideas would simply escape me. I refuse to accept that this is age-related memory loss! So I started using AI as an active reading partner midway through the year. There’s good news and bad news. The good news is that I seem to have found a way to stop my brain from leaking information. AI has turned reading from passive consumption into a constant thinking loop. I’ll share the bad news later, but first, here is the workflow I use to make ideas stick.

Step one: Brain dump, don’t summarise

After a chapter, I don’t worry about elegant notes. I just dump my messy thoughts - what I understood, what confused me, what felt important—straight into the AI.


Prompt: “Outline the latest developments in the DPDP Act in India and its implications for our industry. Summarise perspectives from regulators, in a concise manner.”

Step two: Merge your e-reader notes

I still highlight on my e-reader, but those notes no longer sit in a digital graveyard. I export them and have the AI combine them with the insights from my brain dump to create a cohesive map of the book.

Prompt: “Simulate ‘what-if’ market scenarios for new digital procurement trends, incorporating real-world data. Include risks and counterarguments.”

Step three: Build a ‘concept map’

We all have a pre-existing map of concepts in our heads. New concepts stick better when they hook onto something we already know – it’s called “schema integration” AI is surprisingly good at cross-linking different sets of concepts
Prompt: “Help construct an impact model for rolling out a new digital capability, such as a generative AI-powered knowledge management system for internal teams.”

Step four: Compress and personalise

To wrap things up, I avoid long, useless summaries and ask for something actionable.

Prompt:“Draft a framework for organizing a learning sprint on hyper automation in finance operations. Include evidence-based insights.”

Parminder Singh is cofounder of two AI ventures--ClayboxAI and Kampd--and has held APAC leadership roles at Google and Twitter. For feedback, please email to eteyeonai@timesofindia.com