AI Superiority: From Chess to Table Tennis, Where Machines Have Proven Stronger and Where Humans Remain Unrivaled
In April 2026, the Sony Ace robot with artificial intelligence defeated professional table tennis player Mia Kihara. The match was played according to the official rules of the International Table Tennis Federation (ITTF). Sony's developers call this event historic — a machine has reached an expert level in a real competitive sport for the first time, marking a true breakthrough.
Five iconic AI victories over humans
Against the backdrop of the rapidly developing AI market, reports of robots defeating humans have become more frequent. But it is these five breakthroughs from past years that laid the foundation for the industry. Here are the key milestones:
- 1997 — Chess: Deep Blue defeats Garry Kasparov. The first victory over a champion in a classical match.
- 2011 — Jeopardy! Quiz: IBM Watson crushed the best players, demonstrating AI's ability to recognize complex speech patterns.
- 2016 — Game of Go: AlphaGo defeated Lee Sedol 4:1, overcoming the barrier in a game with an astronomical number of moves.
- 2017 — Poker: Libratus won over $1.7 million in chips. AI learned to bluff under conditions of incomplete information.
- 2019 — Esports (Dota 2): OpenAI Five defeated the world champions — team OG — live on air.
Humans still win
However, machine dominance cannot yet be called an absolute rule. For example, in May 2026, the humanoid robot F.03 from Figure AI lost to an ordinary intern named Aime. Their package sorting showdown lasted 10 hours and was broadcast live.
Each participant had to scan a barcode, lift a box, and place it label-side down on a conveyor belt. The cycle repeated continuously. In the end, Aime processed 12,924 packages, while the machine's result was 12,732 items. This means the human spent 2.79 seconds per object, and the robot 2.83 seconds. Notably, the employee had breaks for rest and meals under California law, while the AI only pulled ahead in the fifth hour when the human stepped away.
In fairness, by the end of the experiment, the intern had developed blisters and his hand was very tired. The robot, however, can work non-stop, so the human's minimal lead over a short distance does not guarantee long-term efficiency. Currently, physical labor allows humans to stay ahead, but for office workers, the situation may change faster.
Additionally, there is an important economic argument. Today, employers widely acknowledge that hiring people is more profitable than maintaining AI. Corporate spending on technology is growing too fast. For this reason, Microsoft is limiting internal licenses for Claude Code for staff due to token costs, and Uber exhausted its entire AI budget for 2026 in four months. Per-minute payment for computing power often eats up all the benefits of workforce optimization.
Expert opinion: Machines win where algorithms are clearly measurable, but in physical work and financial costs, humans still hold the lead. The coming years will show whether AI can overcome these barriers or if the human factor remains the decisive advantage.