The Practice Of Designing Your Information Diet Deliberately

The food diet analogy is imperfect but useful. What it captures:

- Inputs determine capacity. What you consume creates the raw material for thought. You cannot think thoughts that your inputs don't contain the elements for. - Quality and quantity are both variables. You can eat too much of good food and too little of bad. Similarly, volume of information consumption is a separate dimension from quality. - Digestion is required. You don't get nutrition from food you can't digest, and you don't get knowledge from information you never process. - Habit shapes intake. Most of what you eat is habitual, not chosen in the moment. Most of what you consume mentally is the same.

What the analogy misses: information has externalities that food doesn't. What you believe and understand affects how you act in the world, which affects other people. A badly calibrated information diet doesn't just make you mentally unwell — it makes you a less effective participant in every collective system you're part of.

The nutritional science of information is underdeveloped relative to the nutritional science of food. But some things are clear enough to build on.

Most information consumed today is highly processed. Not in a pejorative sense necessarily — processing information to make it accessible is a legitimate and valuable function. But the processing strips things out.

A primary source — a scientific paper, an earnings transcript, a legal brief, a book written by someone who spent years on a topic — contains the author's actual argument, evidence, caveats, uncertainty, and methodology. A derivative source — a summary, an article, a tweet thread about the paper — contains the summarizer's interpretation of that argument, filtered through what they found interesting, what fit their editorial frame, what was quotable, and what their audience would share.

Each transformation step: - Removes nuance and qualification - Selects for interestingness over completeness - Adds the summarizer's interpretive layer - Often misrepresents statistical uncertainty - Usually omits methodology that would allow you to evaluate the claim

By the time you see "Studies show that [surprising claim]," you're often five transformations from the actual research, and the claim has accumulated distortions at each step that make it essentially unreliable as a factual input.

This doesn't mean derivative sources have no value. They serve as discovery mechanisms — they can alert you to things worth investigating directly. The mistake is treating them as the end of the research rather than the beginning.

The most practical framing for redesigning an information diet: start with your questions, then find sources that answer them.

This inverts the default. The default is: here are the sources I habitually check, let me see what's in them today. The intentional approach: here are the things I'm genuinely trying to understand, let me find what addresses them.

Your question set will have layers:

Standing questions — things you're consistently trying to understand that don't change week to week. For a builder: how do markets develop, how do organizations fail, what psychological principles drive user behavior? These are served by long-form material — books, deep research, expert interviews, case studies.

Active project questions — things you need to understand for the specific work you're doing right now. These are more time-bound and specific: what has been tried in this space, who are the domain experts, what does the data show about this specific claim? These are served by targeted research — not ambient consumption.

Current events filter — what from the news stream is actually relevant to your standing questions or active projects? Most of it isn't. The filter should be tight.

When you design from questions, you also discover how little of your habitual consumption actually addresses any of them. Most of it is ambient noise — vaguely interesting, sometimes entertaining, but not connected to anything you're actually trying to figure out. That's not automatically wrong, but it should be chosen rather than defaulted into.

Ungoverned consumption means you consume when there's a pull — when you're bored, when a notification arrives, when a gap appears in your day. This means consumption is reactive and continuous rather than chosen and bounded.

The scheduling principle: treat information consumption like any other production activity. Assign it specific time slots. Have a start and an end. Then stop.

This works for a few reasons:

It makes consumption a choice. When you decide in advance "I'll check the news from 8-8:30 and then again at 6pm," every other moment is automatically a no. You're not making a fresh decision each time a notification arrives. The habit architecture makes consuming outside those times the exception rather than the rule.

It builds in processing time. If you read for 45 minutes and then have 15 minutes of walking or thinking before the next meeting, you have time to integrate what you read. If you consume continuously, each new piece of information displaces the previous before it's connected to anything.

It reveals how much you were consuming out of habit vs. need. Most people who implement scheduled consumption discover they were spending far more time on information grazing than they realized, and they miss almost none of it once it's gone. The FOMO about missing something important is mostly unfounded — genuinely important things reach you through multiple channels.

The research on news specifically supports this. Studies comparing people who check news frequently vs. those who check once or twice a day find no significant difference in factual knowledge or quality of judgment about current events. Frequent checkers are better at knowing what's happening right now, not at understanding it better.

The line between these two states:

Staying informed means having, at any given time, the information you need to operate well — to make good decisions, to have accurate beliefs about things that matter to your work and life, to be a reliable participant in the communities you're part of. It requires active selection, some primary-source engagement, and regular processing.

Being consumed means the information stream has become the primary activity rather than a supporting one. Signs: you feel anxious when you haven't checked for a few hours; you frequently check without a clear reason; you consume and feel neither more informed nor more capable of action; you're often in a state of ambient outrage or anxiety that isn't attached to anything you can actually do anything about.

Being consumed is partly a product of design — platforms are explicitly engineered to maximize time-in-app, which means maximizing the emotional activation that keeps people checking. Outrage, anxiety, novelty, and social comparison are all high-engagement states. An information system optimized for engagement is not optimized for your understanding or wellbeing.

This is not conspiratorial. It's engineering. The incentive is attention. Attention is monetized. Your understanding and mental health are not the product being sold, so they're not what the system is optimized for.

The intentional diet is partly a defense against this. But it's more than defense — it's construction. You're building an information environment that serves your actual goals rather than the platform's goals.

Audit first. For one week, track what you actually consume: every source, how much time, what you got from it. Most people are surprised. Not because they're doing anything dramatic, but because they've never looked.

Identify your actual question set. Write down the five to ten things you're genuinely trying to understand or decide in the next three to six months. Be specific enough that you could evaluate whether a source addresses them.

Match sources to questions. For each question, identify the best primary or high-quality secondary sources that address it. Books by people who spent years on the topic. Original research, selectively read. Expert practitioners. Good longform journalism that cites its sources.

Set consumption windows. Decide when you'll consume information and for how long. Protect those windows from extending. If you decide 45 minutes of reading per day is right, that's what it is.

Create a reading processing ritual. After consuming, take 5-10 minutes to note: what was actually new here? What does it connect to? What questions does it raise? This alone dramatically increases retention and integration.

Prune your subscriptions and feeds aggressively. If something has been in your queue for more than two weeks without being read, it's probably not actually serving your question set. Remove it. The point is not to have access to a huge corpus — it's to actually read and integrate what matters.

Distinguish between discovery and depth. Social media and aggregators are decent discovery tools — they can surface things you wouldn't have found. But discovery is a function you should spend 10% of your consumption time on, not 90%. Once you've discovered something worth engaging with, go engage with it directly.

Information diet is one of those things where the effects are slow to appear and then hard to undo.

A year of consuming primarily processed, reactive, opinion-heavy media and a year of reading primary sources, thinking carefully, and engaging with well-argued long-form work produce dramatically different cognitive states. Not just in what you know — in how you think. The person who spends a year doing the former develops faster reaction times to outrage triggers, higher tolerance for oversimplification, and weaker ability to hold complex information in mind. The person doing the latter develops slower reaction times (in the good sense — they think before reacting), higher tolerance for complexity, and stronger ability to navigate ambiguity.

You become the shape of your inputs over time. This is not metaphorical. The neural patterns that your habitual mental activity strengthens are literally the circuits you'll use to think with. Choose them accordingly.