How to Improve Your General Knowledge in 10 Minutes a Day

The person who knows something about everything is not smarter than you. They run a different system. Here it is, with the research behind every part of it.

The table always leans toward the person with one more detail. That skill is trainable.

There is a moment at every long dinner when the conversation lands on something nobody planned. The fall of Rome. Why the Suez Canal matters. What actually happens in your body when you cannot sleep. And one person at the table has the extra detail, the one that makes everyone stop chewing and lean in.

I used to think those people had been quietly reading since childhood while the rest of us watched television. Then I started paying attention to how they actually behave, and the picture changed. They are not reading more. Several of them read less than I do. What they have is a loop: a small amount of new material every day, a habit of testing themselves on it, and a stubborn refusal to let anything interesting pass without connecting it to something they already know.

That is the whole thing. It is unglamorous and it is learnable.

I have spent the last two years on the tooling side of this problem. I work at a micro-learning startup, which means I read the memory research for work, I test competing apps for work, and I get to see anonymised data on what thousands of people actually do when they try to learn something in five minutes on a train. Some of what I found contradicted what I expected. Most of it contradicted the advice you get when you search this question, which is almost always some version of read more books and watch documentaries.

That advice is not wrong. It is just missing the part that makes it work.

The short answer

To improve your general knowledge, run this loop every day for about ten minutes:

1. Learn one small thing. One explainer, one lesson, one section of an article. Small enough that finishing it is never in question.
2. Close the source and say it back. Out loud, or to a friend, or into your notes app. If you cannot explain it with the page closed, you have not learned it yet.
3. Rotate topics on a weekly schedule. Science, history, money, psychology, technology, culture. Rotation is what turns a pile of facts into breadth.
4. Hook it to something you already know. New knowledge needs somewhere to attach or it slides off.
5. Meet it again after a day, a week, and a month. Spaced review beats cramming at identical total study time.
6. Take the ten minutes from scrolling, not from your life. Reallocated time costs nothing. Added time gets cancelled by week two.

Everything below is why each step works, which topics pay the highest interest, and the three ways people usually break the system in the second week.

One practical note before any of that, because it changes how hard this is. The loop above has an obvious problem: steps one, two and five all require you to organise something on a day when you do not feel like organising anything. That is exactly the work micro-learning apps were built to take off you. They supply the short daily unit, they ask the questions instead of leaving you to test yourself, and the good ones schedule the review so you never have to think about when a fact is due. I work on one of them, NerdSip, and further down I review the handful I actually use, including ours and including where ours falls short.

If you would rather watch the system than read it, we put the whole thing into a short video:

YOUTUBE EMBED → paste this URL on its own line in Medium and press Enter

[embed]

What general knowledge actually is

General knowledge is your working understanding of many fields at once. Science, history, geography, economics, psychology, health, technology, culture. It is what lets you follow a news story without opening four tabs, hold a conversation with someone whose job you know nothing about, and smell nonsense when somebody confidently cites a statistic.

Two words get used interchangeably here and they should not be.

Common knowledge is what almost everyone in a given culture already has. Water boils at 100 degrees Celsius. There are seven continents. It is a floor, not an achievement.

General knowledge is breadth built on purpose. It is common knowledge plus the second layer: not only that inflation exists, but roughly what causes it; not only that antibiotics exist, but why they do nothing for a cold.

The second distinction matters more, and it is the one that quietly ruins most people’s efforts.

Recognition is the feeling of yes, I have heard of that. It arrives instantly and it feels exactly like knowing.

Recall is producing the thing with the source closed.

Recognition is what documentaries, podcasts and saved threads reliably give you. Recall is what a conversation demands. The gap between them is why you can watch six hours about the Roman Empire and still have nothing to say about it at dinner. You did not fail to pay attention. You built the wrong kind of memory.

Why documentaries, podcasts and saved threads do almost nothing

Most of what you learned this week is already leaving. This is normal, and it is fixable.

In 1885 a German psychologist named Hermann Ebbinghaus did something nobody had thought to do: he ran a memory experiment on himself, rigorously, for years. He memorised lists of nonsense syllables, then tested how many survived after twenty minutes, an hour, a day, a week, a month. He plotted the results.

The curve he drew is brutal and it has held up for 140 years. Roughly half of new material is gone within an hour. About two thirds is gone within a day. After a week you are holding a quarter of it, and the quarter you kept is mostly the parts that already connected to something.

Now put that curve next to how most of us consume information. You listen to a ninety-minute podcast at 1.5x while doing the washing up. You are exposed to perhaps forty distinct ideas. You retrieve exactly none of them. Twenty-four hours later, by Ebbinghaus’s arithmetic, you are down to a handful, and because you never tested yourself you have no way of knowing which handful.

This is the part that stings: it does not feel like forgetting. It feels like knowing. Fluent, easy input produces a sensation psychologists call the illusion of competence. The material was clear, therefore you must have learned it. The clarity belonged to the presenter, not to you.

In 2013 a team led by John Dunlosky reviewed ten common study techniques and graded them on how well the evidence actually supported them. Highlighting, rereading and summarising all came back low utility. Two techniques earned the top rating. One was practice testing. The other was distributed practice, which is the technical name for spreading learning out over time.

Those two are rules two and one below. Almost nothing else on the list mattered.

Why the people who know the most learn the fastest

Here is the mechanism that makes general knowledge worth building at all, as opposed to just looking things up.

New information does not enter an empty room. It has to attach to what is already there. Researchers call the existing material prior knowledge, and it behaves like the hooked side of velcro. The more hooks you have, the more the next thing sticks.

The cleanest demonstration of this is a study from 1988 by Donna Recht and Lauren Leslie. They took a group of junior high students and sorted them two ways: by reading ability, and by how much they knew about baseball. Then they had every student read a passage describing half an inning of a baseball game and asked them to recount it.

The good readers who knew nothing about baseball were beaten by the poor readers who did. Knowledge of the subject mattered more than skill at reading. The weak readers who knew the game could tell you what happened; the strong readers who did not know it produced a jumble of names and verbs.

Scale that up and you get the uncomfortable dynamic behind general knowledge. Knowledge compounds like capital. The person who already knows a little about trade routes finds the article about shipping containers easy, so they finish it, so they know more, so the next one is easier still. The person starting from nothing finds the same article heavy, abandons it, and stays where they were. The gap widens without anyone trying.

The good news inside that is the same fact read the other way. Every hook you add makes the next thing cheaper. The first month is the most expensive month you will ever have. It gets easier from there, permanently.

Every new idea has to hook onto an old one. Breadth is the hook supply.

The four rules that do all of the work

Everything worth knowing about building general knowledge collapses into four rules. You could run them with a library card and a shoebox of index cards, and people did exactly that for centuries.

Rule 1: A little every day beats a lot on Sunday

The spacing effect is one of the oldest and most replicated findings in psychology. Take a fixed amount of study time and spread it across several days, and you will remember substantially more than if you spend the same total time in one block.

A 2006 meta-analysis by Nicholas Cepeda and colleagues pooled 254 studies covering more than fourteen thousand participants. Spaced practice beat massed practice by roughly ten percentage points on final recall. Same material. Same total minutes. Different schedule.

The reason is that consolidation happens between sessions, not during them. Sleep does a large part of the work. When you cram, you skip the part where your brain files things.

Practically: ten minutes a day for a week will leave you with more than a two hour session on Sunday. It is also, conveniently, the version you will actually do.

Rule 2: Ask yourself the question instead of rereading the answer

This is the rule people skip, and it is the one that matters most.

In 2006 Henry Roediger and Jeffrey Karpicke ran a study that should be taught in every school. Students read a passage. One group reread it repeatedly. Another group read it once and then took repeated recall tests with no feedback.

Five minutes later, the rereaders were ahead. They felt more confident, too.

One week later the picture had inverted. The group that had tested themselves recalled around 61 percent. The rereaders were down near 40. The rereaders had also predicted they would do better, which is the illusion of competence showing up on cue.

Retrieval is not a way of measuring memory. It is a way of building it. Every time you drag something out of your head, you strengthen the path back to it. A question does more for you than a highlighter ever will.

The practical version is embarrassingly simple. Read the thing. Close it. Say it out loud in your own words. Notice the exact place where you went vague, because that vagueness is the only honest report you will get on what you actually learned.

The whole system lives in this ten seconds: source closed, saying it back.

Rule 3: Follow the thing that actually pulls you

There is a version of self-improvement where you assign yourself the Great Books and grind through them out of duty. It has a poor completion rate and a worse retention rate.

Curiosity is not a soft preference. In 2014 Matthias Gruber, Bernard Gelman and Charan Ranganath put people in a scanner, showed them trivia questions, and asked how curious they were about each answer. Unsurprisingly, people remembered the answers they had wanted more.

The interesting part came next. During the wait between question and answer, the researchers flashed an unrelated face. People in a high-curiosity state remembered those irrelevant faces better too. Curiosity had opened a window, and everything that fell through it stuck, including the things nobody was curious about.

So the strategy is not to force yourself through the topic that sounds impressive. It is to chase the one that genuinely hooks you, then let the drag-along effect carry the neighbouring material. You will end up somewhere you did not plan, which is the point.

Rule 4: Meet the fact again just before it leaves

Rules one and two tell you to space and to test. Rule four is what happens when you combine them intelligently.

Spaced repetition means reviewing a piece of information at increasing intervals, timed for the moment just before you would have forgotten it. Each successful retrieval at that edge flattens the forgetting curve a little further, and the next safe interval gets longer. A day. Then three. Then a week, a month, six months.

The economics are absurd in your favour. A handful of well-timed thirty-second reviews can hold a fact for years. The same total time spent rereading holds it for about a week.

This is the one rule that genuinely benefits from software, because the scheduling is arithmetic and humans are bad at arithmetic about themselves. More on that below.

The ten-minute loop, minute by minute

Here is the whole system as an actual routine. It fits in a coffee break.

Minutes 0 to 5. One lesson, one explainer, one section. Pick by interest, not by duty. Stay inside the week’s topic if you can, but never at the cost of boredom.

Minutes 5 to 7. Close it. Say the idea out loud in one or two sentences, as if to a friend who has never heard of it. Note where you stalled.

Minutes 7 to 8. Hook it. Ask one question: what does this remind me of? Anything counts. A film, your job, something your grandmother said. The hook is doing more work than the fact.

Minutes 8 to 10. Review two or three things from previous days. Not rereading. Questions with the answers hidden.

Then close the app and go about your day. The single most common failure mode is not stopping. People get interested, do forty minutes, feel virtuous, and are gone by Thursday.

The weekly rotation that turns facts into breadth

Daily learning alone produces depth in one direction, which is a fine outcome but not the one you came for. Breadth needs rotation, and rotation needs a schedule, because left to yourself you will read about the same three things forever.

Assign one theme per day:

• Monday: Science. Physics, biology, chemistry, how the scientific method actually works.
• Tuesday: History and geography. Timelines, regions, trade, migration, why borders are where they are.
• Wednesday: Money and systems. Economics, incentives, inflation, markets, how institutions behave.
• Thursday: Psychology and behaviour. Biases, habits, memory, motivation, social influence.
• Friday: Technology and AI. How the tools work, basic technical literacy, what the words mean.
• Saturday: Culture and ideas. Art, philosophy, language, architecture, religion.
• Sunday: Review, then follow whichever thread pulled hardest all week.

Two things happen once this runs for a month. The first is obvious: your coverage evens out. The second is not. Ideas start colliding across days. You read about incentives on Wednesday and it explains something about antibiotic resistance you met on Monday. Those collisions are the actual product. Breadth is not a longer list of facts. It is the number of connections available between them.

One theme a day. The collisions between them are the point.

The eight topics that unlock the most other topics

Not all general knowledge pays the same interest. A few areas act as keys: learn them and dozens of other subjects become readable. If you want the highest return on the first three months, start here.

1. The scientific method. Controls, sample size, peer review, what a study can and cannot show. This is the topic that lets you evaluate every health claim and every piece of technology hype you will meet for the rest of your life.

2. Basic statistics. Averages versus medians, base rates, correlation versus causation, what a percentage is a percentage of. Most public misinformation is not a lie. It is a true number placed next to a misleading frame.

3. World geography. Regions, borders, oceans, trade routes, where things are made and how they move. Geography is the skeleton that news stories hang on.

4. Modern history. Industrialisation through to the internet age. Almost every current argument is a rerun of an older one, and knowing the original makes the current version legible.

5. Economics basics. Incentives, supply and demand, inflation, interest rates, what a market does. You do not need the maths. You need to know why people behave the way they do when money is involved.

6. Human behaviour. Memory, habit formation, cognitive biases, social influence. This one is recursive: it improves how you learn everything else.

7. AI literacy. What a model is, what training data does, why systems hallucinate, how to verify an output. This has moved from optional to load-bearing in about three years.

8. Health basics. Sleep, nutrition, exercise, stress, and how to read a medical claim. The subject with the highest volume of confident nonsense per square inch.

You do not need mastery in any of them. You need enough to ask a decent second question.

Find your gaps before you fill them

Most people improving their general knowledge accidentally study what they are already best at, because it is the most enjoyable material and the least confronting. The fix is one honest baseline.

Take a mixed-topic test once, note which categories embarrass you, and weight your rotation toward those for a month. We built a free twenty-question general knowledge test that scores you by category for exactly this. Any comparable test works. What matters is that the diagnosis comes before the treatment.

Then retake it in three months. This is the only reliable way to see progress, because the subjective experience of getting more knowledgeable is nearly nil. You do not feel yourself improving. You just notice one day that you had something to say.

The best micro-learning apps that I actually use

Twenty minutes of commute is the cheapest learning time you will ever find.

Disclosure before anything else: I work on NerdSip. It is first on this list and I am obviously not a neutral party about it, so I have written down where it loses as well as where it wins, and the other three are apps I genuinely use rather than filler. Treat the whole section with the scepticism it deserves and go try two of them yourself.

None of these apps are necessary. The four rules work on paper. What software buys you is the removal of friction: the scheduling of reviews, the supply of good short material, and the small daily nudge that gets you to open the thing at all. That is worth a lot in practice, because the failure mode is never I do not know how to learn. It is I did not open anything today.

We put several of these side by side on camera, if you want to see them running rather than described:

[embed]

1. NerdSip: the one built around the four rules

Best for: general knowledge specifically, and for people who want the memory system handled for them.

NerdSip opens on a feed of short courses that you swipe through the way you would swipe through anything else on your phone. History, psychology, money, science, the strange corners. Something catches you, you tap in, and you get a deck of short lesson cards you can read or press play and listen to like a podcast.

Two things make it relevant to this article rather than just adjacent to it.

The first is that you can type any topic at all and it generates a full course in about a minute. The engineering inside Roman aqueducts. How anaesthesia works. Why airlines oversell seats. This matters for rule three, because the standard failure of curated apps is that your actual curiosity does not appear in their catalogue.

The second is memory mode, which tracks everything you have learned and resurfaces it as a quick review right before you are due to forget it. That is rule four running automatically, and it is the difference between an app you enjoyed once and knowledge you still have next year.

Quizzes after each course handle rule two. Streaks, XP, leagues and friends handle the showing-up problem, which is a real problem and not a childish one.

Where it loses: the illustration is functional rather than beautiful, and it will not win on visual polish against Imprint. If the aesthetic is what gets you to open an app, that is a real cost. There is also a genuine risk with any AI-generated course that a detail is wrong, which is why we run a fact-checking layer and why you should still treat any single card as a starting point rather than a citation.

Price: free tier is usable. Paid tiers unlock more generation.

2. Imprint: the most beautiful way to learn

Best for: dense subjects that other apps make boring, and for anyone who learns visually.

Imprint is the looker of the group and it is not close. Every lesson is illustrated properly, and the illustrations do work rather than decorate. Big ideas turn into simple swipeable cards, and subjects that normally require effort, behavioural economics, philosophy, the famous books, become genuinely easy to move through.

Its library spans psychology, history, philosophy and book summaries. Short quizzes check that each idea landed, and it builds your feed around topics you pick.

Where it loses: you are inside a curated catalogue. If your curiosity wanders somewhere Imprint has not made a course about, you are out of luck.

3. Yuno: the competitive one

Best for: people who learn better when there is something to win.

Yuno is quiz-first. Each subject is a path of bite-sized quiz levels you climb, so retrieval practice is not an afterthought bolted onto the end of a lesson, it is the entire interaction. It also has full courses, including ones you can just listen to, across history, science and geography.

The best feature is duelling a friend head to head in quick trivia rounds. That sounds like a gimmick and it is not, because it forces recall under mild pressure, which is closer to the dinner-party situation than any solo review will ever get.

Where it loses: quiz-first means the explanations are thinner. You get the fact more reliably than you get the understanding.

4. Deepstash: the book-ideas one

Best for: getting the useful parts of non-fiction without reading forty books.

Deepstash breaks the world’s best books into bite-sized insights and gives you a daily hit of them. It reads you a quick idea, then checks that it really landed, then throws a real-life scenario at you so the lesson has somewhere to attach. Rule two and rule four, wearing a different hat.

Two features I like more than I expected: you can photograph a physical book and it pulls out the key ideas, and everything you save organises itself into stashes rather than dying in a bookmark folder.

Where it loses: it skews heavily toward self-improvement and business thinking. Excellent for how-to-think, thinner on the science, history and geography that general knowledge is mostly made of.

Which one for general knowledge specifically

I am not going to score these out of ten, because the honest answer is that they are built for different jobs and the feature lists move every few months. Pick by what stops you.

If what stops you is not finding the topic you are actually curious about, you want an app that generates rather than curates, which among these four is NerdSip. If what stops you is boredom, take Imprint, because it is the one you will keep opening. If what stops you is not bothering to test yourself, take Yuno, since the entire interaction is a question. If what stops you is wanting the useful part of books without the books, take Deepstash.

The one thing worth checking on any app before you commit, whichever you choose: does it bring old material back to you on its own? Some of these do and some do not, and it is the feature with the largest effect on what you still know in a year. It is also the one nobody puts on the App Store screenshots, so go and look in the settings.

The stack I actually run is NerdSip in the morning for the rotation and the reviews, and Yuno in the evening when a friend is in the mood to lose.

What our own data says about how long to sit there

One more thing, and it is preliminary rather than settled.

We preregistered an analysis of anonymised in-app learning data before we looked at it, mostly to test a folk belief we half-believed ourselves: that accuracy decays steadily the further you get into a session, so short sessions must be better.

It did not replicate. Accuracy is roughly flat across a normal session. The decline people talk about is not there in the first ten or twelve lessons.

Two other things did show up. Past about fifteen lessons in a single sitting, performance does start to fall away, which is a real ceiling and a good argument against the marathon session. And the relationship between speed and accuracy is U-shaped: the very fastest answers and the very slowest answers are both less accurate than the ones in the middle. Answering instantly usually means recognition rather than recall. Answering very slowly usually means you did not have it.

This is one dataset from one app, it has not been peer reviewed, and I would not build a curriculum on it. But it does support the boring recommendation. Ten minutes is not a compromise forced by your schedule. It is close to the size the evidence supports anyway.

Three ways people break this in week two

Collecting is not learning. The most common failure I see is a library of twenty saved courses and zero finished ones. Saving something produces a small hit of the same satisfaction as learning it, at none of the cost. Finish one. Then start the next.

The streak starts serving itself. Streaks work, which is why every app has one. The failure is when a two minute nothing-session to protect a number replaces the actual habit. The streak exists to get you to the learning. When it becomes the thing you are protecting, delete it and keep the habit.

Keeping it to yourself. Knowledge locks in when it leaves your mouth. Tell one person one thing you learned today. It is the single highest-return habit on this page and it costs nothing, because explaining is retrieval practice with stakes.

The arithmetic nobody runs

Ten minutes a day is over sixty hours a year. Nobody feels it happening.

Ten minutes a day sounds like too little to matter. Run the numbers anyway.

Ten minutes a day is 3,650 minutes a year. That is over sixty hours of focused, tested, spaced learning. For comparison, a typical university module is around forty hours of contact time.

One concept a day is 365 new hooks. Not 365 facts, which would be a party trick. 365 attachment points that make everything you read next year slightly easier to absorb, which is the compounding part.

And the time is not new time. Phone-usage reports put the average adult somewhere north of four hours a day on their device. You are not being asked to find ten minutes. You are being asked to move ten minutes from one column to another.

What three months actually feels like

Nothing, for a while. That is the honest answer and it is why most people quit.

Week to week you will notice no change at all, because memory does not announce itself. There is no sensation attached to having a slightly denser map of the world. The improvements are invisible from the inside right up until they are not.

Then somewhere around the third month, somebody mentions supply chains, or the Suez Canal, or why the sky is blue, and you find that you have something to say. Not a recited fact. An actual opinion, with a reason under it. That is the moment the whole thing pays out, and it arrives without warning.

Becoming the most interesting person at the table was never really the goal. It is a side effect of showing up daily for a quarter of a year.

If you want a harder deadline than that, go to a pub quiz. It is the cheapest honest test of whether any of this stuck, because a quiz gives you no time to reconstruct anything and no way to fake recognition. You either have it or you sit there watching it refuse to arrive, which is a genuinely useful thing to feel. Quiz night also punishes exactly the weakness this system fixes: everybody has heard of the answer, almost nobody can produce it. Take a team, lose the first few, notice which rounds you are dead weight in, and weight your rotation toward those. If you want to prepare properly rather than just turn up, we keep a 500-question pub quiz and the round-by-round breakdown on the site.

Quiz night is the cheapest test of whether any of this stuck.

Frequently asked questions

How do you increase your general knowledge?

Learn one small concept a day, recall it from memory before moving on, and rotate topics weekly across science, history, money, psychology, technology and culture. Ten focused minutes daily beats occasional cramming, because memory responds to repetition and retrieval rather than to session length.

How can I improve my general knowledge easily?

Replace ten minutes of scrolling rather than adding ten minutes of study. Use material short enough that finishing it is never in doubt, follow whichever topic genuinely interests you that day, and let an app handle the review scheduling. Effort spent on willpower is effort not spent on learning.

What are the most important general knowledge topics?

The scientific method, basic statistics, world geography, modern history, economics basics, human behaviour, AI literacy and health basics. Each one makes several other subjects easier to understand, which is what makes them worth doing first.

How long does it take to improve your general knowledge?

Expect a noticeable shift at around three months of daily practice. Individual facts stick within days. The sensation of breadth, of having something to contribute across topics, is a compounding effect and it takes a quarter of a year to become obvious.

Can I improve my general knowledge just by reading more?

Partly. Reading supplies the input, but without retrieval you build recognition rather than recall, and recognition collapses under a follow-up question. Reading plus one minute of saying it back is worth several times reading alone.

What is the best app to improve general knowledge?

Choose by how much of the loop the app closes without you. Apps that supply a short daily lesson, quiz you on it, and schedule the review automatically do the most work. NerdSip is built around exactly that loop, Yuno is the strongest quiz-first option, and Imprint is the best if illustration is what keeps you opening an app.

Does general knowledge actually make you smarter?

It makes you faster at learning, which is most of what people mean by smart. Prior knowledge determines how much of any new text you can absorb, so breadth is not a collection of trivia. It is the substrate that everything else attaches to.

Start today, small

Pick tomorrow’s topic now. Set the alarm for a time you are already holding your phone. Learn one thing, close the app, say it out loud, and tell someone at some point during the day.

Do that until it is boring, and then keep doing it, because boring is what a habit feels like from the inside.

If you want the longer version of the system, with the research links and the weekly planner, we keep it all in one place at the NerdSip general knowledge hub. If you want the tool that runs the loop for you, NerdSip is free to start.

Now go learn something interesting today.

I work at NerdSip, which is one of the four apps discussed above. The research cited here is linked at the source and is independent of us. The opinions about competing apps are my own and nobody paid for placement.