A critical review of AI

Man vs. Machine: The Race AI Has Yet to Win

In recent years, AI has achieved remarkable advancements, leading many to believe that the day when it outperforms humans in physical competitions might not be far off. But on a cloudy November morning in Japan’s Suzuka Circuit, it became evident that, for now, AI is still playing catch-up.

This was the premise of a much-anticipated showdown, where a self-driving race car equipped with cutting-edge AI technology squared off against former Formula One driver Daniil Kvyat. Despite the buzz, the AI-driven vehicle didn’t even make it to the starting line—it crashed on its way there.

The Challenge of Racing Without a Driver

The AI-powered car, which had a 90kg onboard computer, was a marvel of innovation. Yet, as it approached a sharp corner enroute to the starting position, it lost traction, spinning out of control. The crash damaged its rear tires and suspension, requiring a tow back to the garage.

The development team attributed the mishap to excessive wheel spin, caused by underinflated and insufficiently warmed tires—a stark reminder of AI’s current inability to adapt to rapidly changing track conditions. Unlike seasoned human drivers, who instinctively sense subtle tire issues or adjust to slipping wheels, AI systems struggle to process such variables in real time.

The Limitations of AI on the Track

The head of the AI team, part of Abu Dhabi’s Technology Innovation Institute, compared the development of autonomous race cars to teaching a toddler to walk: slow, methodical, and prone to falls. The car’s cockpit computer processes over a terabyte of data per minute from its seven cameras, four radars, and numerous sensors. Yet, even with advanced software and hardware improvements, the AI remains several laps behind human adaptability.

Despite this, optimism abounds. The team leader predicts that within a year, these AI-driven cars could match human drivers in speed and precision. In two years, they might even compete against professional racers safely.

A Vision for the Future of Racing

The dream of a future racing league where human drivers and AI machines compete as teammates is already taking shape. Daniil Kvyat, accustomed to the intensity of Formula One, approached the challenge with the same mindset: “I don’t think about who—or what—I’m competing against. I see a challenge and aim to overcome it.”

While the human vs. AI race in Suzuka ended prematurely, other events in the Autonomous Racing League (A2RL) have shown glimpses of what’s possible. Earlier this year, at the Yas Marina Circuit in Abu Dhabi, 12 teams showcased AI-driven cars in the league’s debut. While technical issues arose, such as cars misinterpreting safety protocols and halting mid-race, the event was a step forward in merging AI innovation with motorsport.

Beyond the Track: AI's Broader Purpose

AI racing isn’t just about competition. These experiments push the boundaries of autonomous driving technology, with the ultimate goal of improving real-world applications. Lessons learned on the racetrack could one day make driverless cars safer and more efficient on public roads.

However, not everyone is convinced that autonomous racing will capture the imagination of traditional motorsport fans. Events like this evoke nostalgia for the famous human-machine showdown of Garry Kasparov versus IBM’s Deep Blue in chess. While fascinating at the time, chess enthusiasts ultimately preferred human matches over watching computers compete.

Will AI Ever Truly Replace Humans?

Oxford philosopher Nick Bostrom, renowned for his work on AI’s societal implications, envisions a future where machines outperform humans in nearly every domain. Yet, even in such a future, he argues, there are realms where AI cannot replace human touch—like a child’s drawing for a parent or the thrill of human rivalry in sports.

In an interview, Bostrom speculated that AI-driven sports could coexist with traditional ones but would never fully replace them. “If companies or teams create robots with relatively equal resources to ensure a genuine contest, it could carve out a niche. But most people will still prefer watching humans compete.”

The Road Ahead

For now, AI-driven racing remains in its infancy. It’s a promising yet imperfect spectacle. Suzuka’s audience didn’t witness the exhilarating showdowns of Formula One legends like Ayrton Senna and Alain Prost. Instead, they caught a glimpse of something entirely new—a sport in the making.

The journey from stumbling toddler to sprinting champion is far from over. But if recent advancements are any indication, the day when AI truly rivals humans on the track may be closer than we think.