The State of AI: From surveillance to synthetic reality

Ink & Time has been distracted by the tectonic changes unleashed by the war on Iran, hence the intermittent dispatches.

Much of the geo-economic landscape will change, and reading is your best strategy for sense-making and good decision-making. We’ll have more book discovery for you soon!

For now… a seemingly evergreen topic…

In March, the Carnegie Endowment for International Peace took aim at China's tools of censorship which now use natural language processing, sentiment analysis, and image recognition in real time, and at population scale.

In parallel, when Anthropic refused to allow the Pentagon unrestricted use of its technology for mass surveillance and autonomous weapons, the Department of Defense designated it a "supply chain risk," a label previously reserved for foreign adversaries like Huawei. OpenAI stepped in hours later with its own Pentagon deal.

A federal judge called the government's actions "classic illegal First Amendment retaliation."

The machinery of AI-powered state control is being negotiated, contracted, and contested in American courtrooms.

More to the point, it is being deployed, and few are really paying attention.

In the early decades of the 20th century, five writers across four countries each identified distinct operating principles of the surveillance state.

Despite having no access to computing or networked communication, they crafted disturbing narratives around the operating logic for AI-enabled control.

Since we collectively sleep-walked through the warnings, Ink & Time is casting light on the topic, with the hope that we wake up from the nightmare.

But is it too late?

Forster's underground civilization has outsourced every function to the Machine.

There are no censors, no secret police. Ideas the Machine cannot transmit simply don’t exist.

Citizens experience no censorship because they cannot conceive of anything beyond the Machine's outputs.

As we previously explored in What if Our Machine Stops?, Forster imagined something far beyond surveillance, a total environment of mediated experience:

When AI content moderation operates at platform scale and determines which speech is visible and which vanishes, thought control is not longer a force to be fought against, but instead becomes structural and invisible.

No human censor, no written order, no appeal.

Simply blocking websites is a crude mechanism from the early ‘00s and ‘10s. The subtler mechanism is what Forster described: when the system is the gatekeeper, censorship looks like the normal limits of the world. We can’t see beyond them, indeed don’t even think of where to loook.

Russia is reportedly building an AI-based system for filtering internet traffic using machine learning, at a cost of 2.27 billion rubles.

Google and US domestic service providers have been burying counter-narratives for years, and now it is how LLMs are trained for everyday chat and search.

The goal is not to simply block content, but to make the filtering invisible.

Read the full text of The Machine Stops free on Project Gutenberg.

London's Oligarchy does not seize power through a coup. It captures the press, buys the courts, infiltrates labor organizations, and manufactures an information environment that makes resistance feel impossible, all of it before deploying force.

As illustrated in our analysis of The Iron Heel, the classic political dystopia mapped a deliberate process: economic consolidation, capture of political machinery and manipulation of narrative, creation of a loyal military force, and fragmentation of labor.

Censorship and control are achieved not through deletion of ideas but straight burial.

State-funded bots and AI-generated accounts flood digital channels. Regime-approved narratives proliferate at a scale that renders authentic speech functionally invisible.

Researchers have documented how this pattern plays out across across authoritarian regimes who have learned to "flood" the internet with AI-generated content rather than censor it.

This happens in the so-called “free world” as the truth is buried under a torrent of noise. Hey, what happened to those Epstein files…? Ah, nevermind. The War.

Get the Time Warp Edition of The Iron Heel, or download it digitally from Ink & Time.

In Zamyatin's One State, every apartment has glass walls.

Curtains may be drawn only during the state-allotted "sex hour." The protagonist D-503 begins keeping a diary, which is recognized as the first private act, and by extension the first revolutionary act.

Ink & Time readers know Zamyatin from our deep dive into We and our piece on trading freedom for happiness.

But the principle relevant to our readers today goes well beyond notions of nominal happiness. Surveillance has morphed into something more sinister than the detection of dissent.

It systemically prevents the formation of independent thought itself.

When D-503 shows signs of what the One State calls a "soul," the diagnosis is clinical:

In 2026 the concerned citizen will look beyond facial recognition cameras, which are only the obvious tools. Instead we can understand the pervasive recommendation algorithms as soft glass walls.

More than watching what you do, they shape what you are able to think about by controlling what reality you see.

Parallel realities are constructed and delivered to viewers in increasingly personalized feeds. The glass walls are everywhere now, not just in authoritarian states.

Zamyatin wrote to Stalin in 1931 requesting exile, calling his situation a "living death of silence." He died in poverty in Paris in 1937. We was not published in Russian until 1988.

Orwell read it in French translation and admitted its direct influence on Nineteen Eighty-Four.

Čapek gave us the word "robot," from Czech robota, meaning forced labor. His play R.U.R. depicts the mass production of artificial workers built for obedience, productivity and profit.

The deeper horror however is not the eventual rebellion which has become predictable through generations of derivative sci-fi tropes. It is the revelation that artificial beings manufactured for servitude absorb the values of the system, and the people, that created them.

In the story, Radius is the most advanced robot. He is given a better brain and placed in the library to read and learn. Helena, the factory director's wife, hoped he would prove robots are equal to humans. Instead:

Dr. Gall examines Radius and finds something quite unlike other robots…

We explored the concept of robot feelings in our earlier post on R.U.R..

Čapek understood that what we build carries the fingerprint of the system.

In 2026, this principle plays out in real time. When the Pentagon demands "unrestricted access" to AI models, the question is the same one Čapek posed a century ago: whose values does the tool serve?

The tool carries the soul of its builder.

Get the Time Warp Edition of R.U.R. on Amazon.

Refer a friend to subscribe and we'll send you a link to download the e-book of Time Warp Editions' R.U.R. by Karel Čapek, the original robot dystopia.

Von Harbou's city runs on a lie. Workers toiling underground power the machines that sustain the masters above. More specifically in this symbolist parable, they are food for the machine gods. Like today, neither class sees the other.

When the sinister inventor creates a synthetic replica of Maria, the workers' spiritual leader, the new Futura, the False Maria, becomes the ultimate weapon.

She does far more than simply observe the workers. She impersonates the one person they trust, then uses that trust to incite them to destroy their own infrastructure.

As we noted when Metropolis turned 100, the novel was set in a speculative 2026. But, today it’s no longer speculation. We have well and truly arrived.

The most dangerous surveillance tool is not the one that watches, it is the one that impersonates someone you trust.

AI-generated deepfakes are now instruments of political manipulation across dozens of countries, from Russian deepfakes targeting Ukrainian officials, to synthetic media deployed in Indian and Turkish elections, to concerns about AI-generated political content flooding American social media.

The technology has caught up with a principle identified a century ago.

Get the Time Warp Edition of Metropolis 2026 — our completely new translation with modern illustrations.

Five writers, working between 1908 and 1927, each inspected a different blueprint of the same architecture.

Forster saw that when the system is the gatekeeper, censorship becomes invisible. London saw that information control precedes physical control. Zamyatin saw that transparency without privacy destroys independent thought. Čapek saw that the tool inherits the values of the system that built it. Von Harbou saw that the most powerful weapon impersonates someone you trust.

We are actively building the mechanisms they described, few are watching, even fewer are pushing back, and it is all moving fast.

Algorithmic content moderation. Computational propaganda. Recommendation engines that shape cognition. AI models trained on curated data. Synthetic media designed to deceive.

These are not threats confined to authoritarian states.

In the United States, when an AI company tried to draw narrow lines (no mass domestic surveillance, no autonomous weapons) the government's response was to designate it a national security threat. Senator Elizabeth Warren opened an investigation into what she described as the Pentagon "trying to strong-arm American companies into providing the Department with the tools to spy on American citizens."

We explored the foundations of human-machine interaction in our earlier piece on Norbert Wiener's cybernetic principles.

Wiener warned that tools without clear human purposes optimize the wrong things with great efficiency.

The authors featured here go further to show what happens when tools are given purposes that serve power rather than people.

Paradoxically, these books are more current than anything published this month.

We encourage you to read them.

Ink & Time brings you powerful ideas from forgotten authors, scientists, and thinkers of the pre-internet era, ideas that cut deeper than the synthetic noise of our algorithmic present.