On January 4, 2025, OpenAI CEO Sam Altman posted a simple six word story on social media: “near the singularity; unclear which side.” That brief message captures the most unsettling question in modern technology. As artificial general intelligence inches closer to reality, the lines between human cognition and machine processing are blurring fast. The future of free will is no longer just a philosophical debate discussed in university classrooms. It is an immediate engineering problem that directly affects human dignity.
Five Hundred Million Years of Cognitive Upgrades
Half a billion years ago, the first brains could only perform basic steering functions. A creature either moved toward an energy source or swam away from a predator. Biological evolution took an immense amount of time to build the complex thinking apparatus we use today. Author Max Bennett outlines these crucial leaps in his book A Brief History of Intelligence, showing how each new layer of the brain unlocked an entirely new way of interacting with the physical world.
The progression from simple reactions to abstract thought did not happen overnight. Each stage built a foundation for the next, eventually leading to the deep reasoning capabilities that define human beings.
| Evolutionary Stage | Approximate Timeline | Key Cognitive Feature |
|---|---|---|
| Steering Brain | 500 million years ago | Basic decision making and predator avoidance |
| Learning Brain | 450 million years ago | Memory retention and adaptation through experience |
| Imagining Brain | 320 million years ago | Simulation of possible futures for basic problem solving |
| Reasoning Brain | 70 million years ago | Abstract thought, long term planning, and symbolic logic |
The jump to the imagining brain allowed early species to stop relying purely on trial and error. By simulating possible futures to solve problems internally, animals could avoid physical danger while testing out different strategies. This internal theater of the mind eventually paved the way for complex social cooperation and advanced communication.
The Biological Engine Behind Human Agency
Mammals are entirely unique in possessing a neocortex, which acts as the biological headquarters for our highest cognitive functions. Within that structure lies the prefrontal cortex, the specific region responsible for what philosophers have long called free will. Richard Levy of the Paris Brain Institute points out that without this specific brain region, human decision making would be purely reactive. We would simply respond to immediate environmental cues without any long term planning or moral consideration.
This part of the brain allows us to suppress impulses and deliberately imagine alternative actions before making a move. It is the physical anchor for our agency. However, modern neuroscientists often debate just how independent our choices really are when subjected to subconscious processing.
“You can do what you decide to do, but you cannot decide what you will decide to do.” – Sam Harris, Neuroscientist and Philosopher
That biological limitation becomes especially relevant when we introduce artificial intelligence into our daily routines. If our choices are already shaped by biochemical algorithms, feeding our brains perfectly optimized digital content could easily manipulate our decisions without us ever noticing.
Silicon Systems Reaching the Reasoning Stage
Meta chief AI scientist Yann LeCun recently noted that existing language models do not understand the physical world as well as a housecat. Yet, the technology is advancing through evolutionary stages at a blistering pace. Early machine learning models relied entirely on reinforcement learning, which perfectly mirrors the primitive steering brains of early biological creatures.
Today, the industry is pushing hard toward the final cognitive leap. According to recent discussions on how digital agents still lag behind humans, researchers are actively trying to bridge the gap between pattern recognition and actual independent thought. The frontier of development involves enabling models to perform abstract reasoning and autonomous decision making without continuous human prompting.
We can map current digital progress directly onto our own biological history:
- Early reinforcement algorithms function like a steering brain, learning basic rewards.
- Large language models act like an imagining brain, generating text to simulate possible answers.
- Modern virtual assistants attempt to mimic human social interactions with noticeable limitations.
- Future autonomous agents aim to replicate the reasoning brain through symbolic logic.
If artificial networks successfully replicate this final stage, the definition of intelligence changes permanently. A machine that can plan, reason, and act on its own challenges our exclusive claim to logical agency.
The Mathematical Threat to Independent Thought
Historian Yuval Noah Harari warns that the combination of biological knowledge and computing power creates a dangerous new dynamic. If an algorithm knows your preferences, fears, and habits better than you do, it can predict your choices before you even make them. This concept essentially hacks the human mind by turning our predictable biological responses against us.
Here is a helpful video exploring how technology is fundamentally altering human evolution and power structures:
The financial incentives driving this technology are staggering. The global market size is projected to reach US$827 billion by the year 2030. With that much capital involved, companies are highly motivated to build systems that can predict and manipulate choices through biochemical algorithms to maximize user engagement and corporate profits.
| Public Sentiment and Adoption (2024-2025) | Reported Data Point |
|---|---|
| Americans who believe algorithms harm personal privacy | 53 percent |
| U.S. workers actively using artificial intelligence for daily tasks | 21 percent |
| Predicted unionization increase among knowledge workers by 2028 | 1,000 percent |
The anxiety surrounding these tools is entirely rational. When automated models take over tasks once performed by people, they do not just replace labor; they slowly erode the value of shared effort that builds resilient communities.
Legal Frameworks and the Push for Alignment
The European Union rolled out the AI Act in August 2024 to categorize algorithms by their potential threat level. This sweeping legislative move represents the first comprehensive legal framework categorizing systems by risk. It explicitly prohibits technology that uses subliminal techniques or deceptive tactics to distort human behavior.
Industry leaders are also recognizing the urgency of the situation. In March 2023, the Future of Life Institute published an open letter demanding a six month pause on training models more powerful than GPT-4, citing profound risks to society. Even the organizations building these models understand the stakes. You can review the official charter outlining their commitment to avoiding applications that unduly concentrate power.
“To hack a human being is to get to know that person better than they know themselves. And based on that, to increasingly manipulate you.” – Yuval Noah Harari, Historian
The control problem identified by Nick Bostrom remains unsolved. Ensuring that a superintelligent system remains fully aligned with human autonomy is a monumental technical hurdle. If we fail to solve it, the algorithms we create could easily bypass our prefrontal cortex entirely.
Protecting Shared Effort and Moral Choice
Imagine a digital assistant taking over every complex task that requires careful deliberation and shared effort. Philosophers like Immanuel Kant tied our intrinsic dignity directly to moral agency. If machines make all the optimal choices for us, we lose the daily practice of exercising rational thought.
The authors of the recent book Genesis argue that we must evaluate technological impact against the preservation of dignity. This requires us to ask difficult questions about our future:
- Does automating daily tasks erode the intrinsic value of human labor?
- When algorithms influence our decisions, do they reduce our ability to exercise moral agency?
- If a chatbot simulates empathy perfectly, does it redefine genuine relationship building?
- Can a society maintain its shared values if machines handle all abstract reasoning?
The choices we make over the next few years will dictate the trajectory of human evolution for centuries. As we build systems that mirror the reasoning brain, we have to decide what role biological intelligence will play in an automated world. The tools are already built, the frameworks are being drafted, and the responsibility falls on us to keep our moral agency intact. Navigating this new reality requires us to remain fiercely protective of the very #FreeWillDebate that makes us human, ensuring that #ArtificialIntelligence serves our dignity rather than replacing it.
Frequently Asked Questions
What is the connection between the prefrontal cortex and free will?
The prefrontal cortex is the area of the mammalian brain responsible for long term planning, abstract reasoning, and suppressing immediate impulses. Neuroscientists point to this specific region as the biological foundation that allows humans to deliberate and make independent choices rather than simply reacting to stimuli.
How does the European Union AI Act protect human agency?
The EU AI Act entered into force in August 2024 and categorizes systems by risk levels. It explicitly prohibits algorithms classified as an “unacceptable risk,” which includes systems that use subliminal techniques or manipulative tactics to intentionally distort human behavior and erode independent decision making.
Why did experts call for a pause on AI training in 2023?
In March 2023, the Future of Life Institute published an open letter asking labs to pause training on systems more powerful than GPT-4 for six months. The experts cited profound risks to society, emphasizing the need to develop shared safety protocols and ensure advanced algorithms remain aligned with human values.
What does it mean to “hack” a human being?
Historian Yuval Noah Harari uses this phrase to describe a scenario where algorithms combine vast computing power with biological data to understand a person’s habits better than they do themselves. This allows the system to accurately predict and manipulate the user’s choices, bypassing their conscious free will.
Are current language models capable of true reasoning?
Not yet. According to AI scientists like Yann LeCun, current large language models excel at simulating possibilities and mimicking social interaction, but they lack a true internal world model to understand physics, causality, or independent abstract reasoning.
