Brandolini’s Law in the Age of Artificial Intelligence
Or Why Producing Nonsense Has Never Been So Easy
Maurício Veloso Brant Pinheiro
Keywords: Brandolini’s Law, Brandolini law and artificial intelligence, AI and misinformation, fake news and artificial intelligence, AI hallucinations, economics of misinformation, epistemology of artificial intelligence, AI and truth, influencers and misinformation
There is an informal law that explains the intellectual state of the internet better than many sophisticated sociological theories.
It was not born in universities.
It did not emerge from a laboratory.
And it certainly does not depend on advanced mathematics.
It is simply a brutally honest observation about how public discourse actually works.
It is known as Brandolini’s Law, formulated in 2013 by the Italian programmer Alberto Brandolini.
Brandolini arrived at this conclusion after observing political debates online and noticing something that any attentive internet user quickly recognizes: producing a flawed argument is trivial; dismantling it requires work.
He summarized this phenomenon in a sentence that would become famous:
“The amount of energy needed to refute bullshit is an order of magnitude larger than to produce it.”
The statement seems almost obvious — and perhaps that is precisely why it is so powerful.
Interestingly, the idea is much older than the internet itself.
Centuries earlier, the Irish satirist Jonathan Swift had already captured the same dynamic with literary elegance:
“Falsehood flies, and the truth comes limping after it.”
Between Swift and Brandolini lies a span of three centuries. Yet both point to the same conclusion: lies have always had a logistical advantage over truth.
The internet merely made that advantage impossible to ignore.
The Fundamental Asymmetry of Information
Brandolini’s Law describes a profoundly asymmetric phenomenon.
Producing a false claim requires very little:
- a superficial interpretation
- an incomplete reading
- an exaggerated conviction.
- Sometimes not even that.
In many cases, simply imagining something plausible is enough.
Refuting such a claim, however, requires a far more costly process:
- verifying data
- consulting sources
- analyzing context
- explaining concepts.
While someone can invent a narrative in a matter of seconds, dismantling it may require hours or even days of intellectual work.
This difference in cost creates an interesting phenomenon: lies have a structural advantage in the marketplace of ideas.
Not because they are more convincing.
But because they are cheaper to produce.
The Democratization of Opinion
For centuries, humanity’s central problem was access to information.
Books were rare.
Libraries were limited.
The circulation of knowledge depended on institutions.
The internet solved this problem in spectacular fashion.
Today, anyone has instant access to staggering amounts of information.
But the solution brought an unexpected side effect.
If in the past it was difficult to publish something, today it is difficult not to publish.
Before the internet, producing information required:
- an editor
- review
- printing
- distribution.
Today it requires little more than a smartphone and some confidence in one’s own opinion.
The consequence is predictable.
The internet did not merely democratize information.
It democratized misinformed opinion.
The Problem of Experts
There is also a certain irony in this process.
Experts tend to speak with caution.
They use words such as:
- evidence
- hypothesis
- probability
- uncertainty.
Science, by its very nature, is cautious.
Pseudo-experts, on the other hand, speak with absolute conviction.
They have no doubts.
They have certainties.
And in the environment of social media, confidence often appears more convincing than knowledge.
Thus, a video recorded inside a car can reach millions of views while scientific research remains confined to academic circles.
Not because science is wrong.
But because radical simplification is often more seductive.
The Rise of Instant Influencers
The internet did not merely democratize opinion. It created a new class of informational intermediaries: influencers.
Historically, public knowledge was mediated by imperfect but relatively stable institutions — universities, newspapers, scientific academies. These institutions had obvious flaws, but they possessed at least one basic filtering mechanism: reputation built over time.
The logic of social media replaced that criterion with something far simpler: engagement.
In the new attention economy, authority is no longer built on evidence but on metrics.
- Likes.
- Shares.
- Views.
In this environment, Brandolini’s Law finds its ideal ecosystem.
Producing simplified, provocative, or scandalous content is cheap. Refuting that content is expensive. And while specialists spend hours analyzing data, influencers can produce opinions at industrial scale — often about topics they have never studied.
The asymmetry is evident.
A researcher may spend decades accumulating knowledge in a specific field. An influencer can compress that complex subject into a thirty-second video with dramatic music and a categorical conclusion.
And for a large part of the audience, both messages appear equally legitimate.
Not because they have the same epistemological value.
But because they appear in the same feed.
The consequence is predictable: intellectual authority becomes indistinguishable from performative confidence.
Experts speak cautiously.
Influencers speak with certainty.
And on the internet, certainty usually wins.
In this context, Brandolini’s Law ceases to be merely an observation about the difficulty of refuting nonsense. It becomes an almost perfect model for understanding the architecture of contemporary digital communication.
Never has it been so easy to produce opinion.
And never has it been so difficult to produce knowledge.
Artificial Intelligence Enters the Story
A new protagonist now enters this story: generative artificial intelligence.
Language models and multimodal systems can produce, in seconds:
- complete texts
- plausible articles
- automated comments
- hyper-realistic images
- synthetic videos.
This capability radically changes the cost of content production.
If Brandolini’s Law described an asymmetry between lies and truth, artificial intelligence has dramatically amplified that asymmetry.
Producing plausible content has become almost free.
The Industrial Production of Misinformation
Before artificial intelligence, producing misinformation required human effort.
Today it requires little more than a prompt.
A single operator can generate:
- thousands of texts
- propaganda campaigns
- armies of automated comments.
The verification process, on the other hand, remains essentially human.
- Journalists investigate.
- Scientists analyze data.
- Experts write responses.
One argument at a time.
Lies have gained industrial scale.
Truth continues to be produced by hand.
The Problem of AI Hallucinations
There is also a curious detail: artificial intelligence systems can produce false information without any intention to deceive.
This phenomenon is known as AI hallucination.
Language models are designed to generate plausible sequences of words, not necessarily to verify facts.
As a result, they may produce:
- nonexistent quotations
- invented data
- incorrect references.
These errors do not arise from bad faith.
They arise from the very statistical architecture of these systems.
In other words, artificial intelligence does not merely amplify human misinformation — it can also generate misinformation automatically.
The Paradox of AI Itself
Curiously, the same technology that facilitates the production of misinformation can also help combat it.
Algorithms can detect:
- deepfakes
- bot networks
- logical fallacies in discourse
coordinated propaganda campaigns.
Artificial intelligence thus operates simultaneously as a:
factory of noise
and
filter of noise.
Yet even in this scenario, Brandolini’s Law still holds.
Producing noise remains easier than filtering it.
The Collapse of Digital Evidence
There is an even deeper risk.
When images, videos, and voices can be artificially generated, a troubling phenomenon emerges: the collapse of digital evidence.
If any video can be a deepfake, then any evidence can be contested.
A real video can be dismissed as fake.
A fake video can appear real.
In such a scenario, the truth no longer needs to be refuted.
It only needs to be made debatable.
And when everything becomes debatable, public trust begins to erode.
An Old Philosophical Lesson
Brandolini’s Law is not merely an observation about the internet.
It reveals something far older about human nature.
Thinking requires effort.
Believing does not.
Perhaps that is why the philosopher Friedrich Nietzsche observed that convictions are more dangerous enemies of truth than lies.
And perhaps that is why the psychologist Daniel Kahneman showed that the human brain tends to prefer quick answers over analytical thinking.
Stupidity, it seems, is energetically efficient.
The Economics of Lies
If we analyze the phenomenon from an economic perspective, Brandolini’s Law describes a profound shift in the production of information.
Lies have a low marginal cost.
Truth has a high verification cost.
Artificial intelligence has widened this gap even further.
Producing misleading content has become cheap.
Producing reliable knowledge remains expensive.
In economic terms, this is a deeply asymmetric market.
Conclusion: Truth as a Handcrafted Product
Brandolini’s Law has become one of the most important observations for understanding the informational ecosystem of the twenty-first century.
Artificial intelligence did not create the problem.
It merely amplified it.
Today, lying on a global scale has become trivial.
Truth, on the other hand, still requires what it has always required:
- time
- evidence
- critical thinking.
In other words, truth remains a handcrafted product in a world that produces nonsense on an industrial scale.
And perhaps this is the final irony of the digital age.
Never have we had so much access to knowledge.
And never has it been so easy to ignore it.
#ArtificialIntelligence #AI #BrandolinisLaw #Misinformation #Disinformation #FakeNews #TechPhilosophy #PhilosophyOfTechnology #CriticalThinking #AIethics
Copyright 2026 AI-Talks.org
