I read that whole book and I could not tell you one thing from it
Read a whole book and can't recall a single point? Here's a low-friction way to capture what hit you and find it later in plain language.
I Read That Whole Book and I Could not Tell You One Thing From It
You finished a book three weeks ago. You loved it. You told two people about it. You posted a story with the cover. You highlighted half of chapter four. Then last night, at dinner, someone asked, "Wait, what was that book about?" and your brain produced exactly nothing. A blank, faintly embarrassed nothing. "It was really good," you said. "It was about… I think communication? Or attention? Something about attention." The person nodded politely. You went home and felt slightly hollow about it for the rest of the evening. (Same hollow feeling as the book a friend recommended three months ago you still haven't bought.)
This happens with everything now. Books, podcasts, articles, documentaries, that incredible Substack post you read on the train, the Andrew Huberman episode about sleep, the Ezra Klein interview that genuinely changed how you think about politics for about forty-eight hours. You consumed it. You felt the click. You meant to do something with it. Now, three weeks later, the only residue is a vague feeling of "that was good" with none of the actual content underneath. (Information was never the problem: retention was.)
It's not just embarrassing at dinner. It's the slow horror of realizing that hundreds of hours of your reading life are just... gone. Not stored. Not retrievable. Felt and forgotten.
Why does reading not stick the way we want?
This isn't a personal failing. It's how memory works. Reading is a high-bandwidth, low-encoding activity. You take in thousands of words at a pace far faster than your brain can convert into long-term memory. Research on memory going back to Ebbinghaus has long suggested that new information fades fast unless you actively revisit it, with most of it directionally slipping away within days.
The problem is worse with audio and video. You're listening to a podcast while doing dishes, you nod along with a brilliant point, the dishes are done, the point is gone. You watched a 90-minute documentary about food systems and could probably tell someone the vibe of it but not a single specific claim, statistic, or quote.
We also read more than ever, and across more sources. The number of books, articles, podcasts, and videos a single person consumes in a year is genuinely huge, and our brains were not designed to retain that volume. Without an external system, retention is going to be miserable. Not because you're bad at reading, but because biology says so.
What systems do people try, and how do they break?
The first attempt is highlighting. You highlight half of every book you read on your Kindle, you star half the lines in every article in your reader app, you bookmark every podcast episode that hits. The act feels productive. It is, in fact, doing almost nothing. Decades of studies on highlighting consistently show that passive highlighting has near-zero effect on retention. You're marking, not processing.
The second attempt is a notes app. Apple Notes, Notion, a Moleskine, Roam, Obsidian. You start strong on a book, write a few thoughts, abandon by chapter five, and never look at the notes again. The notes become a pile of half-formed thoughts that you couldn't search even if you wanted to. Same trap that kills every productivity system you've tried.
The third is the readlater pile. You'll write up your thoughts on this book later, when you have time. Later does not arrive. The pile grows. The pile becomes shame. This is why read-later apps end up unread, and why Udemy courses sit at 12% complete.
None of these are character flaws. They're systems where the friction of saving correctly is high and the payoff is delayed by months. Of course you stop.
What does retention actually require?
The research on what makes ideas stick is annoyingly clear. You need to process material, not just receive it. Saying it back in your own words, even briefly, is more effective than highlighting fifty passages. You also need to be able to retrieve it later, and retrieval works on cues, on the questions you'd ask in real life, not on filing conventions you set up at the time.
In practice that means two things. First, the second a sentence hits you, you need to capture it in your own language, not the author's, even just a five-second voice note saying "the thing about how attention is the rarest form of generosity." Second, when you want it back six months later at dinner, you need to find it by asking "what was that thing about attention being a kind of generosity," and have the system actually understand the question.
The books-and-articles industry has built tools for the first half (highlights, bookmarks, clippings) and almost nothing for the second. You can find your highlights from a specific book if you remember which book it was. You cannot find "the thing about attention" across everything you've ever read. That second half is the missing piece. (Podcast episodes saved have the same gap.)
How does dEssence make what you read findable?
dEssence is memory you don't have to maintain. When a sentence in a book hits you, you capture it from whichever surface is closest: the Chrome extension, the Telegram bot, or the web app at dessence.ai. The capture is closer to texting yourself than to building a filing system. Save it, forget it, ask for it later. You ask in your own words. No folders, no tags, no organizing.
For reading, the workflow is brutally simple and that's the whole point. When a sentence in a book hits you, you take a screenshot of it on your Kindle and save it, optionally with a one-line voice note saying what it meant to you. "This is the thing about attention being generosity, blew my mind, want to use this." Done. Same for an article: clip it with the Chrome extension, add a sentence. Same for a podcast: voice note while you're walking the dog, "Huberman just said the thing about morning sunlight calibrating cortisol, look this up."
Later, when someone asks at dinner, you pull out your phone and type "that thing about attention as generosity" or "the morning sunlight cortisol point." It comes back: your screenshot, your note, the source. You actually have the idea, in retrievable form, instead of a vague feeling.
Because the search works on meaning, not tags, you don't need to have filed anything as "productivity" or "psychology" or "podcast notes." You just dumped them in. Searching the way your brain naturally remembers, the gist of the idea, not the source citation, finds them. Which is what you actually want.
What changes when reading is actually retrievable?
Reading goes from a leaky bucket to something cumulative. You stop feeling vaguely embarrassed when people ask about the books on your nightstand. You start using ideas in conversation, in your own writing, in arguments, because they're available, not buried. The two hours you spent on a book stop being two hours of vibes and become two hours of usable content.
It also changes what you read. You become more willing to read difficult, slower stuff because you trust you'll be able to come back to specific passages. You become less of a completist and more of a skimmer-with-precision: read the whole thing once, save the four passages that mattered, move on.
It stacks. A year of reading-with-capture is a real personal library, searchable in your own words. A year of reading-without-capture is, unfortunately, a year of mostly nothing. You feel that difference in the third year. Same way one dropped follow-up becomes the rule.
Frequently asked questions
Isn't this just note-taking with extra steps?
No. Note-taking apps require you to organize, tag, and revisit. dEssence requires you to dump and forget. The retrieval is the part that's different: you ask in your own words and it finds the thing. There's no folder structure to maintain.
Do I have to take notes on the whole book?
No, and you shouldn't. Save only the four or five passages that genuinely hit you. Quality over coverage. A handful of well-captured ideas per book beats a thousand passive highlights.
What about audiobooks and podcasts where I can't highlight?
Voice notes. While you're listening, hit pause, talk into the bot for 10 seconds: "the part about how grief comes in waves, that On Being episode with Gregory Orr on shaping grief with language." That's enough to find it later.
Will I actually use it, or will this just become another forgotten app?
The key is that it lives where you already are: a browser extension, a chat thread, a web tab. Saving is one tap. That's why it tends to stick where notes apps don't. The friction is closer to zero.
Can it pull together notes across multiple books on the same topic?
Yes. Search "everything I've read about sleep" or "all the parenting stuff I've saved" and it surfaces relevant captures across books, articles, and podcasts. That's where it stops being a notes app and starts being an actual personal memory.
A note on the big-knowledge-base trap
There's been a wave of content over the last few years about building elaborate personal knowledge systems in Notion, Roam, Obsidian, where you're supposed to capture every idea you encounter and link it to every other idea in some grand personal knowledge graph. If that works for you, great. Users on forums and Reddit threads keep describing a familiar arc, though: a weekend setting up the system, a few weeks of use, then the system goes quiet. The setup is heavy, and the payoff arrives only after you've already done a lot of librarian work.
The alternative isn't a smarter graph. It's a much lower bar for capture and a much smarter retrieval step. You don't need to link your Atomic Habits notes to your Stoicism notes via a tag taxonomy. You just need to be able to ask "what was that thing about identity-based habits" in six months and have it come back. The connections are made at retrieval time, by your question, not at capture time, by you doing the librarian work.
This is what dEssence is doing differently. The capture is closer to texting yourself than to building a knowledge base. The retrieval is closer to asking a friend than to running a search query. That gap, easy in, easy out, is the reason this approach tends to stick where the heavy systems don't. What you're building isn't structured. It's just yours, in your own words, findable.
Where dEssence falls short of those tools: dEssence is in beta, the paid tier isn't finalized, there's no native iOS or Android app yet (only the Chrome extension, Telegram bot, and web app), and there are no team or shared lists. If you want a deeply linked graph view of your notes, Notion or Obsidian will give you that. dEssence trades the graph for friction-free capture and meaning-based recall.
Read in a way that adds up
You are reading too much for biology to handle on its own. That's not a moral failing, it's just true. The choice isn't between reading less and forgetting everything. The choice is between consuming with a leaky bucket or consuming with a place to actually keep the parts that mattered.
A screenshot, a one-line voice note, and the trust that you can ask later the way you'd ask a friend. That's the entire system. The next time someone asks what your book is about, you don't blank, you say the thing that hit you, in your own words, because it's still there. The reading you've already done starts paying off. The reading you do next starts compounding.