Hi Bharath, please read our policy on LLM writing before making future posts consisting almost entirely of LLM-written content.
Sometime in early 2024, I stopped reading books. Not because I don’t love the process—but because I realized the alpha/hour ratio was far too low.
Reading texts (especially documentation) has become increasingly inefficient. It’s not about laziness; it’s about leverage. A static wall of text demands effort for diminishing insight. But a PDF piped into an LLM? That’s different. That’s programmable knowledge. I can chat with it. Extract patterns. Mine underdiscussed takeaways. Highlight blind spots. Summarize arguments. It’s like outsourcing the heavy intellectual lifting (Subjective).
This flipped a mental switch for me:
Documents today aren't messages to be read—they're messengers to be conversed with.
If I send someone a PDF, I’m not expecting them to read it cover to cover. I’m expecting them to throw it into their favorite assistant and have a dialogue with my intent. A document, in this sense, is just a vessel—a digital PA that carries my signal forward.
I don’t have a solution yet, but I think it’s obvious:
And sharing won’t look like “sending a file”—it’ll look like:
“Send Bharath’s repo docs to your Claude instance.”
“Put this paper in your weekend calendar as a chat object.”
“Ask my assistant what key takeaways your agent missed.”
Documentation will exist less as content and more as contextual proxies—something to interrogate, not ingest.
We're not building static libraries. We’re building living conversations.
Let that sink in.