The Thinking Game’ Review: DeepMind Documentary Glazes Over AI’s Existential Risks

A new documentary chronicling the rise of DeepMind and its founder, Demis Hassabis, has arrived on YouTube. While The Thinking Game offers an unprecedented look inside one of the world’s most secretive AI labs, it ultimately serves as a polished mirror of Silicon Valley’s greatest blind spot: the refusal to deeply engage with the societal consequences of the technology it is building.

Directed by Greg Kohs, the film follows the trajectory of Hassabis from a child chess prodigy to a Nobel Prize-winning scientist leading Google’s charge toward Artificial General Intelligence (AGI). Yet, beneath the soaring score and tales of scientific conquest, the documentary inadvertently highlights a divide between the technical brilliance of AI creators and their philosophical detachment from the tools they are unleashing.

The “Gamer” Who Wanted to Solve Science

The narrative arc of The Thinking Game is rooted in the “Great Man” theory of history. It paints Hassabis not merely as a CEO, but as a visionary destined to decode the human mind. The film dedicates significant runtime to his backstory, specifically his disillusionment with the world of competitive games.

At age 12, Hassabis was the second-highest-rated chess player in the world for his age group. The documentary retells a pivotal epiphany during a match against a veteran grandmaster. Disgusted by the psychological tricks and wasted intellect involved in pushing wooden pieces around a board, young Hassabis realized the futility of the exercise.

“Is this the best use of all this brain power?” Hassabis recalls asking himself. “If you could somehow plug in those 300 brains into a system, you might be able to solve cancer.”

This moment serves as the foundational myth of DeepMind: the transition from playing games to solving “The Thinking Game” of intelligence itself. It frames the company’s subsequent achievements—mastering Go, Starcraft, and eventually protein folding—as the logical evolution of a child’s desire to make the world more efficient.

When AlphaStar Scared the Strategists

The documentary is at its most compelling—and inadvertent—when it showcases the sheer, alien capability of DeepMind’s creations. A central sequence focuses on AlphaStar, an AI agent trained to play the real-time strategy game StarCraft II.

Unlike Chess or Go, StarCraft involves “imperfect information” (fog of war), bluffing, and long-term economic planning, making it a closer proxy for real-world military conflict. When AlphaStar dismantled professional human players, it didn’t just win; it played with a relentless, machine-like aggression that unnerved observers.

This segment prompts the film’s only significant brush with skepticism. Margaret Levi, a noted political scientist, is shown expressing deep concern regarding the dual-use nature of such technology.

• The Concern: If an AI can outmaneuver humans in a complex war simulation, what happens when it is integrated into autonomous weapons systems?

• The Validation: Former Google CEO Eric Schmidt appears, nodding to the potential for “wars that occur faster than humans can comprehend” and enhanced global surveillance.

For a brief moment, the film peers into the abyss. It suggests that the quest for AGI might produce entities that are uncontrollable.

The “Neutrality” Trap

However, just as the narrative approaches these jagged edges, it retreats into safety. The documentary counters Levi’s fears with a standard Silicon Valley defense mechanism: the argument of technological neutrality.

Hassabis responds with a sentiment familiar to anyone who has debated tech ethics over the last two decades. “Technology is neutral in itself,” he argues. The outcome, he suggests, depends entirely on how society, governments, and companies choose to deploy it.

Critics argue this stance is increasingly insufficient. In an era where algorithms drive polarization and autonomous systems are being tested in warfare, the idea that a tool maker bears no responsibility for the tool’s nature is being challenged by ethicists and regulators alike. By treating AI as a blank slate, The Thinking Game absolves its creators of the need to embed safety rails into the very architecture of their systems.

A Google PR Exercise?

Viewers must also contend with the provenance of the film. While focused on Hassabis, DeepMind is a subsidiary of Alphabet (Google). The documentary, currently available for free on YouTube, arrives at a time when the tech giant is battling for supremacy against OpenAI and Microsoft.

The film acts as a soft-power asset. It humanizes the black-box algorithms of Gemini and AlphaFold by tethering them to the likable, earnest persona of Hassabis. He appears not as a corporate titan, but as a genius striving to cure disease and map biology.

Consequently, the film feels airbrushed. It glosses over the internal corporate struggles, the ethical boards that were dissolved, and the sheer commercial pressure to monetize AGI. It sells the “wonder” of AI to a public that is increasingly anxious about job displacement and digital hallucinations.

Background: The DeepMind Journey

To understand the significance of the documentary, one must understand the unique position DeepMind holds in the tech ecosystem.

The Origins Founded in London in 2010 by Demis Hassabis, Shane Legg, and Mustafa Suleyman, DeepMind began with a mission statement that sounded like science fiction: “Solve intelligence, and then use that to solve everything else.”

The Acquisition Google acquired the lab in 2014 for over $500 million, a move that signaled the start of the modern AI arms race. DeepMind operated with semi-autonomy, keeping its research based in London, distinct from Google’s California HQ.

The Breakthroughs

• AlphaGo (2016): The first program to defeat a professional human Go player, a feat experts thought was a decade away.

• AlphaFold (2020): Perhaps their most significant contribution to humanity, this AI system solved the “protein folding problem,” predicting the 3D shapes of proteins. This has revolutionized biology and drug discovery.

The Goal: AGI Unlike “Narrow AI” (which plays chess or recommends movies), DeepMind’s goal is Artificial General Intelligence (AGI)—a system that possesses the ability to understand, learn, and apply knowledge across a wide variety of tasks, matching or exceeding human cognition.

What Happens Next: The Race for Sovereignty

The Thinking Game ends on a note of anticipation, but the reality on the ground is shifting rapidly.

1. The End of the “Game” Phase DeepMind is no longer just a research lab playing board games. Under the new consolidated “Google DeepMind” brand, the pressure is on to deliver commercial products that compete with ChatGPT. The era of pure research is likely fading in favor of product deployment.

2. The 2026 Regulatory Clash As AGI capabilities improve, the “neutrality” argument put forth by Hassabis in the film will face legal tests. The EU AI Act and pending U.S. regulations will force companies to take responsibility for their models’ outputs. The “it’s how you use it” defense may no longer hold up in court.

3. The Military Question Despite the film’s dismissal of military concerns, the Pentagon and global defense agencies are aggressively courting Silicon Valley. The strategies honed by AlphaStar are indeed being analyzed for real-world tactical applications, regardless of the creator’s intent.

Conclusion

The Thinking Game is a beautifully produced, inspiring look at one man’s intellectual journey. It successfully conveys the brilliance of Demis Hassabis and the monumental achievement of AlphaFold. However, as a piece of journalism or documentary filmmaking, it suffers from tunnel vision.

By treating the social and political ramifications of AGI as an afterthought, the film embodies the very hubris critics fear. It presents the end of human intellectual dominance as a “good thinking game”—a puzzle to be solved—rather than a transformative event that requires extreme caution. For the scientists, it is a game; for the rest of the world, the stakes are significantly higher.