Google’s AlphaGo computer bested a professional player in a number of matches last month, demonstrating such explosive progress in the capabilities of manmade machines.
Trust, however, that humans can still outperform computers when it comes to problem solving — such as when you turn quantum mechanics into gaming, as recently discovered by Danish scientists.
Physicist Jacob Sherson of Aarhus University in Denmark, while confronted with the challenges of creating a quantum computer, heard of the concept of gamification or how difficult problems are crowdsourced through letting the public play games to produce the best solutions. In 2012, he heeded this idea and came up with Quantum Moves, an online game for collaborative volunteer research on quantum physics.
And the problem at hand couldn’t be any more complicated: quantum mechanics.
Unlike the usual computers with bits representing 0s and 1s, a quantum computer stores data in qubits, which can be in 0 and 1 states at the same time, or what is dubbed a superposition of states. Quantum information needs to be protected from all forms of external noise in the surroundings, with the slightest bit of interference causing the system to lose the superposition and errors in calculations to exist.
Quantum Moves helps address the need for quantum computers to executive operations rapidly. Players try to figure out the quantum speed limit or how fast data can be manipulated, which today’s fastest computer algorithms still attempt to calculate and beat.
“Features that used to be uniquely human are fully captured by contemporary algorithms. [But] our results are here to demonstrate that there is still a difference between the abilities of a man and a machine,” explains Sherson, whose team has identified humans’ intuitive thinking and problem-solving approach as an edge over computers’ processing power.
The researchers mapped out how the human brain makes decisions based on experience and intuition using Quantum Moves played by more than 10,000 individuals. Using simple strategies, players demonstrated an ability to break the quantum speed limit through completing 500,000 games.
The game’s original version, called Bring Home Water, required collecting and moving virtual atoms in sitting quantum buckets from a spot to another as quickly as they can. The challenge is atoms don’t like to be stationary, “sloshing” around in the game wells — players then need to move them quickly with minimal sloshing in order to discover optimal solutions.
The researchers visualized it as liquid gathered at a low point of a flexible line, where players are tasked to flex the line such that as much liquid would gather at a target site. When users move it too slowly, they get a poor score; when they move it too quickly, sloshing would occur and also affect their score.
The liquid represents the atom’s quantum state — moving it too rapidly brings excitations in its wave function.
The results: some players beat the best algorithms, willingly sacrificing some stability to experience a good game. These led the researchers to seed their optimization software with the best solutions as described by users, with their analysis producing a value for the quantum speed limit that was less than 70 percent of the value created by the algorithm alone.
This translated to over 30 percent of the time needed on the arduous quantum physical task.
“[O]ur results demonstrate that the comparison between man and machine still sometimes favors us. We are very far from computers with human-type cognition," says Sherson, who has hired professional game designers since the game’s earliest version to provide an improved experience, especially for players with little scientific background.
The team has added another game on the website and plan to extend the approach to an array of other problems.
The findings were published in the journal Nature.