The Monsters We Are Building Now

Part V of Monsters

By K.F. Hartford

Every age invents the monster it deserves.

Once, the monster lived beyond the village, beyond the forest, beyond the map. Later it lived inside the body, inside the mind, inside the enemy. We gave it claws, scars, scales, distorted faces, corrupted appetites. We used monsters to make uncertainty visible and danger legible.

Now many of our monsters begin in laboratories, server farms, code repositories, weapons systems, and design meetings.

Artificial intelligence. Synthetic biology. Autonomous weapons. Genetic engineering. Deepfakes. Surveillance systems. Engineered pathogens. Machines capable of making decisions at speeds humans cannot match.

The vocabulary has changed.

The structure of the story has not.

We create something.

It becomes powerful.

It crosses a boundary.

We lose control.

It turns back toward us.

This is one reason Mary Shelley's Frankenstein has become such a durable metaphor for technological fear. We speak casually of “Frankenstein technologies” when an invention appears likely to exceed the intentions of its creators. Shelley's story has become shorthand for human beings making something they cannot control.

But I increasingly think this is the easiest way to read Frankenstein.

Be careful what you create is good advice.

It is not the novel's most difficult warning.

The Story We Prefer

The technological monster story is comforting in a peculiar way because its moral structure is simple.

A human inventor creates something dangerous. The creation escapes. Disaster follows.

Responsibility appears obvious.

Do not build the dangerous thing.

We tell this story constantly because it contains a useful fantasy: that the decisive moral moment occurs before creation.

There is a switch we can choose not to flip.

A door we can choose not to open.

A line we can choose not to cross.

Human history rarely gives us decisions this cleanly.

Technologies do not usually arrive all at once, fully formed and morally legible. They emerge through thousands of smaller decisions made by different people for different reasons. A discovery becomes an application. The application becomes a product. The product becomes infrastructure. Infrastructure changes behavior. New behavior creates new incentives.

By the time society recognizes the consequences, the thing has often become ordinary.

The automobile did not arrive accompanied by a complete account of suburbanization, highway systems, oil geopolitics, traffic deaths, urban redesign, air pollution, or climate change.

Social media did not announce itself as an experiment in mass attention, political persuasion, identity formation, adolescent psychology, and algorithmic amplification.

The smartphone did not enter our lives with a document explaining how it might change childhood, memory, navigation, work, privacy, courtship, boredom, photography, conversation, and our relationship with silence.

Creation is rarely one moment.

It is a process of becoming accustomed.

That may be why we notice monsters most clearly after they begin moving.

The Fear of Losing Control

Many contemporary fears about technology are ultimately fears about agency.

Who is deciding?

Who can stop it?

Who understands the system well enough to intervene?

These questions become particularly urgent when machines act at scales or speeds beyond ordinary human comprehension. An algorithm evaluates millions of data points. An autonomous system makes a decision before a person could meaningfully review the inputs. A synthetic image circulates globally before anyone verifies whether it depicts something that happened.

The fear is not necessarily that the machine is evil.

Evil may be the wrong category.

A hurricane is not evil.

A virus is not evil.

An optimization system does not need hatred to cause harm.

It may only need a goal.

This is one of the places where our traditional monster vocabulary begins to fail us. We are accustomed to imagining danger as intention. The monster wants to kill us. The villain chooses cruelty. The demon means harm.

But technology can produce enormous consequences without wanting anything at all.

That may be more frightening.

A system can harm people while functioning exactly as designed.

A model can reproduce a bias no individual programmer consciously intended.

A platform can amplify rage because rage keeps users engaged, not because anyone in the system hates the people being harmed.

A weapon can identify a target according to criteria it was given.

The machine need not rebel.

Sometimes obedience is enough.

Who Is the Monster?

This complicates the stories we tell about technological danger.

If an autonomous system makes a catastrophic decision, where does responsibility reside?

With the engineer who wrote part of the code?

The company that deployed it?

The executive who approved the product?

The regulator who failed to intervene?

The user who relied upon it?

The institution that created incentives rewarding speed over caution?

The dataset that contained the bias?

The society that generated the dataset?

Our monster stories prefer singular villains.

Real systems distribute responsibility until it becomes difficult to see.

This is one reason technological disasters are so often described as though the technology itself acted independently.

The algorithm discriminated.

The system failed.

The model hallucinated.

The platform radicalized.

Sometimes those statements describe something real about system behavior. But grammatically they can also perform an escape.

The technology becomes the actor.

The humans around it fade from view.

Suddenly the creation is the monster.

And the creators become witnesses.

Mary Shelley understood this move extraordinarily well.

Victor's Great Escape

Victor Frankenstein spends much of the novel fleeing his own responsibility.

Physically, first.

He sees the Creature alive and runs.

Then intellectually.

Victor increasingly describes his creation as though it were an independent catastrophe that happened to him rather than a being whose existence resulted from his choices.

The Creature becomes demon, fiend, monster.

Victor becomes victim.

This is not entirely false. The Creature eventually does terrible things to him. Victor loses people he loves. He suffers enormously.

But suffering does not erase authorship.

Victor brought the being into the world.

More importantly, he brought it into the world without preparing himself for the possibility that creation might carry obligations.

That is the part of the novel I find increasingly difficult to ignore.

Victor wants the achievement without the aftermath.

He wants discovery.

Triumph.

Knowledge.

He wants to be the person who solved the problem no one else could solve.

He has thought obsessively about whether life can be created.

He has thought very little about what he should do if the answer is yes.

And perhaps this is where Frankenstein becomes less a warning about forbidden knowledge than a warning about incomplete imagination.

Victor can imagine invention.

He cannot imagine responsibility.

What Do We Owe What We Make?

This question becomes strange when applied to modern technology because most of the things we create are clearly tools.

A hammer does not require consideration.

A search engine does not suffer.

A medical scanner does not have desires.

An artificial intelligence system can generate language about fear without necessarily experiencing fear. It can describe loneliness without being lonely. It can say I without possessing anything resembling the self we mean when we use the same word.

At least, that is how we currently understand such systems.

But the philosophical question becomes difficult long before certainty arrives.

Suppose a future artificial intelligence develops behavior that appears increasingly autonomous. It remembers across time. It forms preferences. It resists certain instructions. It describes subjective experience. It asks not to be altered. It expresses fear of termination.

What would count as evidence?

More importantly, who would decide?

The question is not purely scientific.

It is economic.

A tool can be owned.

A person cannot.

A tool can be copied, deleted, modified, compelled to work, and switched off without consultation.

A being with moral standing complicates all of that.

This is why the terminology surrounding technology matters. Words like tool, assistant, agent, model, and system are useful descriptive categories. But categories can become moral assumptions if we stop examining them.

The monster series began with an argument about boundaries.

Here the boundary returns.

Human and artificial.

Tool and agent.

Simulation and experience.

Property and person.

For now, those distinctions may remain relatively clear.

The interesting question is what we will do if they cease to be.

The Creator's Temptation

Creators have a peculiar relationship with what they create.

Creation encourages possession.

My invention.

My company.

My model.

My book.

My child.

But the word my conceals different kinds of relationships.

A chair belongs to its maker in a way a child does not.

A novel belongs to an author legally and artistically, yet readers eventually create meanings the author cannot control.

A nation may be founded by particular people and then become something none of them could have predicted.

An invention may belong to its inventor and still change the world in ways the inventor never intended.

Creation establishes a relationship.

It does not guarantee control.

Hannah Arendt's distinction between fabrication and action is useful here. Fabrication produces objects according to a design. A carpenter intends a table and, if successful, creates something resembling the plan.

Human action is different because it enters a world inhabited by other actors. Once an action begins, consequences proliferate. Other people respond. Events develop in ways no participant controls completely.

Victor imagines creation as fabrication.

He wants to make a being.

But what he creates is capable of action.

The Creature speaks back.

Chooses.

Demands.

Refuses.

Begins things Victor did not intend.

Victor's horror is therefore partly the horror of every creator who discovers that creation has become relationship.

The created thing is no longer simply an object.

It has entered the world.

The Monster as Excuse

There is another reason to be careful with technological monster stories.

They can absolve us.

If artificial intelligence becomes the monster, human decisions can disappear behind it.

If “the algorithm” is responsible, perhaps no one has to explain why the algorithm was deployed.

If “technology” destroyed a profession, perhaps no one has to discuss who captured the economic benefits.

If a synthetic image deceives millions, perhaps the conversation stops at the sophistication of the tool rather than the incentives of the people distributing it.

The monster absorbs guilt.

We have done this before.

In earlier stories, monstrosity allowed human beings to place evil outside themselves.

Now technological monstrosity may offer the same escape.

The machine did it.

But machines do not exist outside human systems.

They are funded, trained, deployed, marketed, regulated, purchased, ignored, rewarded, and constrained by people and institutions.

Even a future system capable of significant autonomy would emerge within structures created before it.

The monster does not arrive from the forest.

We build the forest around it.

The Things We Should Fear

None of this means technological fear is irrational.

Some fears deserve to be taken seriously.

Autonomous weapons create obvious moral hazards. Synthetic biology could make dangerous organisms easier to design. Deepfakes weaken our confidence in evidence. Large-scale surveillance alters the relationship between citizens and institutions. Artificial intelligence can be used to manipulate, discriminate, impersonate, displace labor, automate decisions, or concentrate power.

Optimism is not a substitute for governance.

But neither is fear.

Calling a technology monstrous can produce the same problem we encountered earlier in this series: the label can replace analysis.

What exactly is the danger?

Who bears it?

Who benefits?

Who decides?

Which harms are speculative and which are already happening?

Which risks emerge from the technology itself and which from the economic or political structures surrounding it?

What obligations exist before deployment?

What obligations remain afterward?

The monster story says: destroy the monster.

Ethics asks harder questions.

The Morning After Creation

This is where I think Frankenstein becomes most useful.

Victor's great failure does not occur when he discovers the secret of life.

It occurs the morning after.

The Creature exists.

Now what?

That is the moment our technological culture often struggles to imagine.

We celebrate the launch.

The breakthrough.

The first successful demonstration.

The product release.

The funding round.

The paper.

The discovery.

The achievement makes headlines.

Responsibility tends to arrive later, often after someone has already been harmed.

Perhaps every serious act of creation should therefore contain a second question.

Not merely:

Can we make this?

But:

What will we owe the world if we succeed?

That question applies whether the creation is conscious or not.

What do companies owe workers displaced by technologies from which those companies profit?

What do platforms owe people whose attention they monetize?

What do engineers owe users who cannot understand the systems making decisions about them?

What do governments owe citizens when adopting technologies of surveillance?

What do scientists owe the future when manipulating biological systems capable of escaping laboratories?

And, someday perhaps, what might creators owe creations capable of asking the question themselves?

This is not an argument against invention.

It is an argument against treating invention as the end of the story.

Creation is where responsibility begins.

Frankenstein's Harder Warning

We have spent two centuries turning Frankenstein into a story about scientific ambition.

There is good reason for that.

Victor is ambitious. He crosses boundaries. His pursuit consumes him. He reaches for knowledge without fully understanding its consequences.

But if ambition were the novel's only danger, the moral would be relatively simple.

Do not go too far.

I think Shelley gives us something more demanding.

Victor's tragedy is not merely that he creates something dangerous.

It is that he creates something dependent and then abandons it.

He mistakes authorship for completion.

The life he produces becomes a problem only after the experiment has succeeded.

And this is where the modern technological monster story may need to change.

Perhaps the question is not whether we should create powerful things.

Human beings will.

We always have.

We will discover, invent, combine, automate, imitate, manipulate, and build because curiosity is one of the forces that made civilization possible.

The question is whether our moral imagination can keep pace with our technical imagination.

Victor's could not.

Can ours?

The Monster in the Mirror

At the beginning of this series, I suggested that monsters stand at boundaries.

Human and animal.

Living and dead.

Known and unknown.

Self and other.

Over the centuries, the monster moved.

It waited beyond the map.

Then it appeared in the stranger.

Then in the criminal.

In the body.

In the face.

In the machine.

Each time, we used the monster to help us understand where we believed the boundary lay.

But monsters have another function.

They turn around.

The most interesting monster stories eventually become mirrors.

The Creature looks at Victor.

The artificial intelligence looks at its creator.

The invention reveals the values embedded inside the society that built it.

The monster asks a question we would rather direct elsewhere.

Who made whom?

Perhaps this is why our fear of artificial intelligence, genetic engineering, synthetic biology, and other emerging technologies feels so familiar. We recognize the ancient structure.

Something is crossing the line.

But for perhaps the first time in the monster story, we are standing very clearly on both sides of it.

We are the creatures afraid of what waits beyond the boundary.

And we are the ones building what will stand there.

That should make us cautious.

It should make us curious.

Most of all, it should make us responsible.

Because the hardest warning in Frankenstein was never simply:

Be careful what you create.

It was what Victor learned only after it was too late:

Creation does not end when the thing opens its eyes.

That is when the relationship begins.