The Attention Triage World

In a world saturated with information, the question is no longer how to stay informed, but how to avoid becoming cognitively bankrupt. Marc Andreessen captured this shift in a recent tweet describing his current media diet -- one quarter X, one quarter long-form podcast interviews with practitioners, conversations with AI models, and old books -- adding that "the opportunity cost of anything else is far too high, and rising daily."
The important signal in that tweet isn't the mix itself. It's what's missing. Marc isn't describing a media diet so much as a filtering philosophy. The real shift forced by AI abundance isn't that we should consume different information, but that we should consume far less of it -- far more deliberately.
We are moving from an information scarcity world, where being informed meant reading more, to an attention triage world, where being informed means cutting harder.
1. From "Read More" to "Cut Harder"
For most of modern history, being informed meant increasing inputs. More newspapers, more blogs, more feeds. Information was scarce, so the edge came from access.
That world is over.
AI doesn't just increase information. It destroys the marginal value of most of it. News is instantly summarized. Takes are infinitely replicated. Expertise becomes cheap to query. In that environment, content becomes abundant and judgment becomes scarce.
The question is no longer "How do I keep up?" It's "Which inputs compound judgment instead of noise?"
2. What an Attention Triage World Actually Means
In medicine, triage means deciding who gets attention and who doesn't -- fast. Not everything gets treated. Not everything deserves time.
Now apply that to media.
You wake up to millions of posts, thousands of AI-written summaries, infinite takes, breaking alerts, coordinated outrage, and synthetic authority packaged beautifully. You cannot process it. You cannot even meaningfully sample it.
So the skill that matters is no longer consumption. It is ruthless exclusion.
In an attention triage world, ignoring 95% of inputs is not laziness. It is discipline. It is survival.
3. Why Marc's Split Works
Andreessen's mix works because each quadrant serves a distinct cognitive function:
1. X (weak-signal layer). Not for depth -- for detection. Inflection points. Coordination. Vibe shifts. This layer should be skimmed, not absorbed.
2. High-bandwidth human signal. Long-form conversations with operators, builders, investors, scientists -- people with skin in the game. This is where incentives surface, second-order effects appear, and reality intrudes on theory. As AI floods the world with plausible content, this category becomes more valuable.
3. AI as intellectual exoskeleton. Not for facts, but for synthesis, adversarial questioning, stress-testing beliefs, compressing weeks of reading into structured mental maps. The people who stay informed won't read more. They'll interrogate models better.
4. Time-resistant inputs. Old books, foundational essays, history, philosophy -- the kind of knowledge favored by the Lindy Effect. These inputs don't decay with the news cycle. They calibrate you across cycles and inoculate you against narrative whiplash.
That's not eclectic. It's optimized for compounding judgment.
4. The Deeper Shift: From Consumption to Construction
The best-informed people I know don't "follow the news." They build internal models, update selectively, and ignore almost everything else.
Information now behaves like calories in an environment of hyper-abundance. Most people will overconsume. Some will binge junk. A minority will adopt disciplined diets. An even smaller group will fast aggressively and regain cognitive health.
The uncomfortable truth is that most people will rely on AI summaries of feeds they already like, mistake coherence for understanding, feel informed while becoming more passive, and outsource judgment without realizing it. They'll consume processed narratives, not raw signal. That's fine for orientation. It's dangerous for agency.
5. Who Actually Stays Informed
The winners in the attention triage world will:
- Treat attention as capital.
- Read less but ask sharper questions.
- Use AI to challenge beliefs, not confirm them.
- Prioritize insight over immediacy.
- Understand that ignorance about 95% of things is a feature, not a bug.
In an attention triage world, ignorance isn't the risk. Undisciplined attention is. And the gap between those two is about to define who thinks clearly -- and who merely feels informed -- for the next decade.