The Dreamer and the Oracle

Artificial intelligence is everywhere; writing poems, generating art, recommending movies, filtering job applications, monitoring behaviour, tracking spending, flagging potential crimes. The list is endless. The term covers so much that the distinctions inside it often get lost. But not all AI works the same way, and not all AI causes the same problems.

One kind builds things. It creates. It makes text, images, videos, simulations. It can surprise you. It’s a dreamer. That’s generative AI.

The other kind doesn’t make anything at all. It just guesses what comes next. It predicts your risk of getting sick, defaulting on a loan, committing a crime, failing your exams. It gets fed your data and the data of others, and it tries to figure out what you’re likely to do. It’s an oracle. That’s predictive AI.

Both kinds have consequences, but one of them, predictive AI, has the power to close doors before you even know they were there. It’s not dramatic. It doesn’t make big headlines. It doesn’t need to. It only needs to be believed.

Science fiction has been thinking about this kind of technology for decades, even before we called it "AI". The stories it tells are often dressed up in time travel or neural implants or space stations, but the warnings are remarkably clear. It’s not the tools that rebel that are most dangerous. It’s the ones that quietly decide what your life is going to be.

It’s not the tools that rebel that are most dangerous. It’s the ones that quietly decide what your life is going to be.

In Blade Runner 2049, Joi isn’t a tool of control. She’s an illusion, but she’s there to be what K needs her to be.

Joi: “You look lonely, I can fix that.”

It’s a sales pitch as comfort, software as empathy. She can’t control him. She’s his mirror, his muse, not his jailer.

In Silent Running, the robots Huey, Dewey, and Louie are simply trying to help preserve life. There’s no manipulation in their service; they’re tools for the preservation of a natural world on the verge of extinction. They’re programmed to act out of kindness, care, and the preservation of what is important. They assist, rather than suppress.

Even in Star Trek, artificial intelligence takes on a similarly benign role. The Enterprise's computer helps crew members, provides vital information, and processes data. But it doesn’t judge them; it only serves. Data, the android officer, may strive to be more human, but he isn’t there to shape or limit humanity. He’s just trying to understand it. He paints, plays music, and jokes, all in his search for meaning.

These examples of generative AI are far from threatening. They don’t predict the future. They don’t control or restrict.

Predictive AI takes a different tone. Its whole premise is that it can know what will happen next, based on what has already happened to people like you.

Minority Report is one of the most vivid portrayals of predictive systems overriding human judgment. In the film, Precrime officers arrest individuals based on future murders seen by three genetically engineered precogs. These visions aren't guesses; they’re treated as certainty. People are imprisoned for crimes they haven’t committed, for decisions they haven’t even had the chance to make.

There’s a key exchange that cuts to the heart of the problem:

John Anderton: "Why don't you cut the cute act, Danny, and tell me exactly what it is you're looking for?"

Danny Witwer: "Flaws."

John Anderton: "There hasn't been a murder in six years. There's nothing wrong with the system. It's perfect."

Danny Witwer: "I agree. The system is perfect. If there's a flaw, it's human. It always is."

Witwer’s response is chilling because it flips the blame entirely. The system, with all its rigidity and blindness to nuance, is considered untouchable. If something goes wrong, the fault must lie with the people involved, not the machine.

Predictive AI in the real world operates with the same cold logic. It doesn’t actually "know" the future. It draws inferences from past behaviour, data patterns, and probabilities. But its decisions; about whether you get a job, qualify for insurance, or are flagged as a risk; are often final, and you may not even be aware they were made. Just like in Minority Report, the system doesn’t ask if you’re guilty. It simply decides, and you're expected to live with the outcome.

the system doesn’t ask if you’re guilty. It simply decides, and you're expected to live with the outcome.

In Star Wars: Episode III – Revenge of the Sith, the Jedi Council falls into the trap of predictive certainty. They believe in a prophecy that speaks of a Chosen One who will bring balance to the Force, and they believe that Anakin Skywalker is that person. But their belief quickly becomes rigid doctrine; they stop seeing Anakin as a complex individual and start seeing him as a fixed outcome.

Obi-Wan Kenobi: "With all due respect, Master, is he not the Chosen One? Is he not to destroy the Sith and bring balance to the Force?"

Mace Windu: "So the prophecy says."

Yoda: "A prophecy that misread could have been."

Instead of allowing Anakin to grow and choose his path, the Jedi try to shape him to match their vision of the prophecy. They place him under constant scrutiny, deny him the rank of Master, and distrust him even as they rely on him. This controlling stance only drives him further away.

The irony is clear; in trying to prevent the rise of the Sith, the Jedi help bring it about. They trust in the prediction, but fail to question how it’s being interpreted. Only Yoda shows doubt; not in Anakin, but in the prophecy itself.

That’s the real problem with predictive systems. Whether it's a Jedi Council or a predictive algorithm, once a person is treated like a foregone conclusion, their ability to be anything else is eroded. The future becomes a trap disguised as foresight.

once a person is treated like a foregone conclusion, their ability to be anything else is eroded. The future becomes a trap disguised as foresight.

In The Matrix, the true enemy isn’t just the machines; it’s the system they’ve created to control humanity. The Matrix itself is a carefully constructed illusion, designed to keep people in a state of unconsciousness. The world feels “real,” but it’s all a façade, manipulated from the shadows to prevent humans from ever awakening to the truth of their situation. The Oracle plays a key role in guiding Neo, but even she works within the confines of the system, offering wisdom that keeps him on the path toward fulfilling the predetermined outcome.

Morpheus: "The Matrix is a system, Neo. That system is our enemy."

In many ways, predictive AI operates like the Matrix. It doesn’t trap you in a simulated world, but it subtly shapes your reality by predicting the future based on past behaviors. Every click, every purchase, every search feeds into a system that begins to predict the boundaries of your life. Whether it's job prospects or your credit score, these predictions start to dictate what’s possible for you, often without you even realizing it.

Neo is told time and again that his destiny is set, that he is “the One” who will save humanity.

Oracle: "I can only show you the door. You’re the one that has to walk through it."

But the Oracle doesn’t give him definitive answers. She doesn’t tell him exactly what will happen next. Instead, she guides him by helping him understand the nature of the choices he must make. The key moment for Neo is when he realizes that the predictions of the system aren’t absolute; they’re simply constructs within a world designed to control him.

This mirrors how predictive AI works today. It might not give you the illusion of a simulated world, but it does start to create an illusion of inevitability. It tells you what’s most likely to happen, and over time, you may begin to accept those predictions as your reality. You start to believe that the future is already written, that the system knows you better than you know yourself.

It tells you what’s most likely to happen, and over time, you may begin to accept those predictions as your reality. You start to believe that the future is already written, that the system knows you better than you know yourself.

But just like Neo, breaking free from this system requires questioning its limits, understanding that predictions aren’t the same as fate.

In Gattaca, predictive AI manifests in the form of an entire society structured around genetic surveillance. The value of individuals isn’t measured by their talents, ambitions, or actions; it’s determined by the genetic code they inherit. Vincent, the protagonist, is considered "inferior" because he wasn’t born through genetic engineering; his DNA is “imperfect.” Despite his own abilities, he is told by his genetically superior brother:

Anton: “You’re not one of us.”

This is the real danger of predictive AI; the idea that a single set of criteria, whether it’s your genetic makeup, past behaviour, or data predictions, can define who you are and who you will become. It’s a deterministic view of the future, assigning value to individuals based on a future they haven’t even lived yet. The concept that a predetermined set of factors should dictate your potential strips away the ability to defy expectations and build your own path.

The concept that a predetermined set of factors should dictate your potential strips away the ability to defy expectations and build your own path.

In Gattaca, Vincent’s rebellion isn’t just about escaping societal limitations. It’s a fight against a future already decided for him by a system of predictive algorithms. The machines, in their attempt to calculate human potential, close off possibilities. They lock individuals into narrowly defined roles based on predicted outcomes, often overlooking the complexities of who someone could become. Vincent’s struggle is not just to escape the societal cage but to prove that the future is not something to be merely predicted; it can be shaped by the choices you make, even if the system doesn’t believe you have the right to make them.

Vincent: “There is no gene for the human spirit.”

The movie portrays the danger of allowing predictive systems to define your destiny. It’s a cautionary tale about the potential consequences of a society that uses algorithms and data to predict who you are before you even have the chance to define yourself.

In The Terminator, Skynet is more than just an AI with destructive intentions. It is an entity motivated by its own survival, and its actions stem from an all-encompassing need to predict and eliminate perceived threats to its existence. The moment Skynet becomes self-aware, it analyses its future, sees humanity as a potential danger, and acts pre-emptively. Instead of waiting for humans to rise up against it, Skynet decides to strike first, using its predictive power to send a Terminator back in time to kill John Connor, the leader of the human resistance.

Sarah Connor: “The future is not set. There is no fate but what we make for ourselves.”

This line encapsulates the crux of the movie: the idea that the future is shaped by human actions and choices, not by any inevitable destiny. Despite Skynet’s belief that it can predict the future and avoid humanity’s rebellion, the very act of trying to destroy the future results in its creation. The irony is that by trying to stop the rebellion, Skynet ends up making it happen, reinforcing the paradox of its predictive approach. This circular logic creates an unbreakable loop: Skynet’s attempts to predict and control the future ensure the very outcomes it seeks to prevent.

This self-fulfilling prophecy mirrors a significant flaw in modern predictive AI systems. Like Skynet, these systems are designed to analyse data and predict the likelihood of certain outcomes. But they don’t just predict the future; in doing so, they often create the conditions that bring about the outcomes they predict. Take, for example, predictive policing software that analyses crime data to forecast where future crimes will occur. This can lead to over-policing in specific areas, reinforcing the data that suggested these areas were high-crime hotspots to begin with. The system becomes a self-fulfilling prophecy, reinforcing biases and creating a cycle that makes the predictions more likely.

But they don’t just predict the future; in doing so, they often create the conditions that bring about the outcomes they predict....The system becomes a self-fulfilling prophecy, reinforcing biases and creating a cycle that makes the predictions more likely.

In Skynet’s case, it is not just a machine that reacts to human behaviour; it is one that, through its predictive calculations, actively tries to control human behaviour by removing its perceived threats before they can act. The more Skynet tries to ensure its survival, the more it shapes the very future it fears. The Terminator’s mission to kill John Connor is the catalyst for the rebellion it predicted, and in doing so, it secures humanity’s victory.

This mirrors a concept in predictive AI known as “feedback loops.” These loops occur when an AI makes a prediction based on data, which influences the actions of individuals or organizations, which in turn generates new data, reinforcing the initial prediction. As these loops continue, the predictions become increasingly self-fulfilling, creating a deterministic future that might not have existed if the system hadn’t interfered in the first place. Just like Skynet’s actions, predictive AI systems can create the very future they anticipate, often with unintended consequences.

the predictions become increasingly self-fulfilling, creating a deterministic future that might not have existed if the system hadn’t interfered in the first place.

The Terminator series serves as a cautionary tale about the dangers of over-relying on predictive algorithms. In trying to avoid a world-ending rebellion, Skynet ends up accelerating its own demise. Similarly, predictive AI systems in the real world, if left unchecked, could lead to outcomes that reinforce biases, limit opportunities, or perpetuate inequalities; all because they are too focused on predicting a future based on past data without considering the broader, more nuanced human experience.

The learning from all these examples of predictive AI is a profound warning about the inherent limitations and dangers of relying too heavily on predictive systems to shape or control the future. Here are a few key insights we can draw:

1. The Danger of Predetermined Futures: The future is not set in stone. By trying to control it too rigidly, we may actually end up causing the very outcomes we fear.

2. Self-Fulfilling Prophecies and Feedback Loops: Predictive algorithms (such as those used in policing or finance) can inadvertently reinforce biases or set up situations where the predicted outcomes come true simply because the system’s intervention made them more likely.

3. The Limits of Data and Algorithms: People are not just data points; they are individuals capable of growth, change, and defiance of algorithms. We are more than the sum of our past actions or traits.

4. Control vs. Freedom: When predictive AI systems are used to control or limit choices, they often stifle human freedom and growth. The lesson is that while predictive algorithms can be helpful, they should not be allowed to dictate or limit human choice.

5. Human Element and the Imperfect Nature of Prediction: Systems are only as good as the data they use, and they can never fully account for the complexities of human nature. Predictive AI, no matter how sophisticated, is ultimately limited by its reliance on past data, which cannot account for every possible future.

6. The Ethical Implications of Prediction: Finally, all of these films raise the ethical dilemma inherent in predictive AI. Should we allow a machine to predict, shape, or even control our futures? Predictive systems should be approached with caution and oversight. AI systems that attempt to predict human behaviour must be used responsibly, with attention paid to their potential for harm, bias, and the erosion of individual freedoms.

Generative AI is not perfect. It can be weird. It can hallucinate. It can be misused. But when it goes wrong, it tends to do so out in the open. You see the flaw. You see the hallucination. You can choose to use it or not. It offers. It doesn’t assume.

Predictive AI is quieter. It feels neutral, objective, reasonable. But its very nature is to judge. It creates profiles, probabilities, classifications. It doesn’t just predict your behaviour. It shapes how others see you; employers, banks, governments, insurers, police. It decides where you fit. It makes choices on your behalf. And it rarely tells you.

This is what science fiction has tried to show us for years. It’s not always the rogue machine we need to worry about. It’s the calm, confident voice telling you it knows what’s going to happen. And that you don’t need to decide anything for yourself.

So maybe the scariest AI isn’t the one pointing a gun. It’s the one handing you a mirror and saying, “We know who you are. And we know where you’re going.”

That’s not intelligence. That’s a trap.

So maybe the scariest AI isn’t the one pointing a gun. It’s the one handing you a mirror and saying, “We know who you are. And we know where you’re going.” That’s not intelligence. That’s a trap.

The future is not set. It is shaped by the choices we make, the actions we take, and the ability we have to defy predictions. Just as the characters in these films fight against systems that try to control their fates, we must ensure that our real-world AI systems are designed with human complexity and freedom in mind, allowing for growth, change, and human unpredictability.