Why LindyFact Exists

The Problem

We are drowning in information and starving for knowledge.

Every day, thousands of news articles compete for your attention. Most will be forgotten by tomorrow. The 24-hour news cycle optimizes for what's new, not what's true or important. Breaking news breaks you—fragmenting attention, spiking cortisol, and leaving you less informed than before.

This isn't a bug. It's the business model.

Three Thinkers Who Saw It Coming

Neil Postman: The Entertainment Trap (1985)

In Amusing Ourselves to Death, Neil Postman warned that we wouldn't be oppressed by what we fear, but by what we desire:

“What Orwell feared were those who would ban books. What Huxley feared was that there would be no reason to ban a book, for there would be no one who wanted to read one...”

Postman identified what he called disinformation—not lies, but something more insidious:

“Television is creating a species of information that might properly be called disinformation...”

His diagnosis: The medium shapes the message. When news must compete for attention against entertainment, news becomes entertainment. Substance drowns in spectacle.

Yuval Noah Harari: The Information Problem (2024)

In Nexus, Harari argues that human failure isn't about flawed nature—it's about flawed information systems:

“The problem is not in our nature. The problem is in our information...”

Why doesn't information quality improve over time? Because fiction is cheap and truth is expensive:

“It's very easy to create fictional information because you don't need to research anything...”

His diagnosis: We don't need smarter people. We need better filters. Institutions that “invest in truth” by separating signal from noise.

Nassim Nicholas Taleb: The Lindy Effect (2012)

In Antifragile, Taleb formalized an ancient intuition: time is the only honest judge.

“If a book has been in print for forty years, I can expect it to be in print for another forty years...”

Taleb asks: Who judges the expert? Who guards the guard? His answer:

“Well, survival will.”

His diagnosis: Don't try to predict what's important. Wait and see what survives.

The Synthesis: Why Time-Filtering Works

These three thinkers converge on a single insight:

Thinker	Problem	Solution
Postman	The medium distorts	Change the medium
Harari	Information is cheap	Build filtering institutions
Taleb	Experts fail	Let time be the judge

LindyFact is an attempt to implement all three:

We change the medium by ranking stories by survival time.
We build a filtering institution that “invests in truth” by letting time do the checking.
We let survival be the judge.

What This Means in Practice

What We Show

✓Stories that have survived at least 24 hours
✓Stories with multiple independent sources
✓Stabilizing coverage
✓Lasting entities

What We Filter Out

✗Breaking news (too early)
✗Single-source stories
✗Rapid decay curves
✗“Celebrity slams X” headlines

The Algorithm

Every story receives a Lindy Score $L(a) \in [0, 100]$ based on three weighted components. We publish our algorithm because transparency builds trust.

Input Variables

t= age in hours since publication

c₀= initial citation count (when first seen)

c= current citation count

d= number of unique domains covering the story

t_last= hours since last seen in news feeds

Component Functions

35 pts max

1. Survival Age

Logarithmic formula rewards early survival with diminishing returns over time.

S(t) = \min\left(35, \left\lfloor 10 \cdot \log_2\left(\frac{t}{24} + 1\right) \right\rfloor\right) \cdot \mathbf{1}_{t \geq 24}

A 1-day old story earns ~10 pts. A 7-day survivor earns ~30 pts.

40 pts max

2. Velocity Shape

Measures citation retention. Stories that maintain coverage score higher.

V(c, c_0) = \begin{cases} 40 & \text{if } \frac{c}{c_0} \geq 0.5 \\ 0 & \text{if } \frac{c}{c_0} < 0.1 \\ \left\lfloor 100 \cdot \left(\frac{c}{c_0} - 0.1\right) \right\rfloor & \text{otherwise} \end{cases}

Retaining 50%+ of initial citations = full points. Rapid decay (<10%) = zero.

25 pts max

3. Source Breadth

Rewards stories covered by multiple independent sources.

B(d) = \min\left(25, \left\lfloor 8 \cdot \log_2(d) \right\rfloor\right)

2 sources = 8 pts. 4 sources = 16 pts. 8+ sources = 24 pts.

Zombie Penalty

Stories that disappear from news feeds are penalized. If a story hasn't been cited in 24+ hours, it may be “zombie” content—still indexed but no longer relevant.

Z(t_{\text{last}}) = \begin{cases} 0.8 & \text{if } t_{\text{last}} > 24 \\ 1.0 & \text{otherwise} \end{cases}

Total Score

For Breaking stories (less than 24 hours old):

L(a) = 0 \text{ (not yet scored)}

For Non-Breaking stories (24+ hours old):

L(a) = Z(t_{\text{last}}) \cdot \min\left(100, S(t) + V(c, c_0) + B(d)\right)

Classification Thresholds

Following Taleb's Fragile → Robust → Antifragile framework:

Score	Status	Meaning
t < 24h	Breaking	Not yet tested by time
L > 80	Antifragile	Gains strength from time
60 ≤ L ≤ 80	Lindy	Robust—survives time
55 ≤ L < 60	Uncertain	Borderline longevity
L < 55	Anti-Lindy	Fades with time—noise

The logarithmic terms reward early survival with diminishing returns, while linear velocity interpolation penalizes rapid citation decay.

The Core Thesis

The news that matters will find you—if you wait.

LindyFact is a bet: that time-filtering will surface better information than recency-filtering.

A Note on Humility

We are not claiming to have solved misinformation. We are not fact-checkers. We are not arbiters of truth.

We are building a filter based on a simple hypothesis: survival is a signal.

Stories that survive multiple news cycles have passed a test. Not a perfect test, but one harder to game than clicks.