XMaze

Crowd-sourcing Solutions to Really Big Problems

Imagine you want to solve a big problem X (place anything instead). For a concrete example, take human aging problem. The money is not a bottleneck. To solve it, you have to analyze the mechanisms of aging. For that you need people who understand the mechanism well enough to actually do useful computational and experimental work.

Much of it is computational, such as applying machine learning algorithms to do manipulate the data generated, - to be able to recognize a feature of a molecular structure, a pattern of asteroid distribution, or the possibilities in a network of chemical reactions of a human cell.

Here is where the 3D Study Maze video game comes in. It is a 3D labyrinth game, where each room's walls are covered in theory to be learned, and each door in a tunnel between the rooms contain a problem based on the theory in the room: to unlock the door, you have to solve the problem; and, as you move along to other rooms, you may need to use the theory of the previous rooms.

While entertaining kids, the game teaches them starting from the very basic mathematical knowledge and skills (integrating the Khan Academy videos and problem sets), to advanced university-level knowledge (integrating freely available material of on-line courses).

The game creates a motivational element: game players have to solve problems to go from one virtual room another. Curiosity is known to be sufficient to create game addicts across the world, but to increase the performance kids, and appease parents, we could offer extra monetary rewards for solving harder problems.

The educational problems would be only in the beginning of the game. Once the kids arrive to the place where they have skills necessary to do useful job, educational problems gradually are replaced by real life problems, placed in a similar way like advertisements are through Google AdWords program.

The type of problems a player is capable of efficiently solving will depend on the set of maze rooms that the kids were able to reach so far. This information will enable us to target real world business and academic problems to the right place of the educational maze, where the likelihood for a type of problem be solved is optimal, from perspective of the value of having the problem solved.

Each place in the labyrinth, of course, will have different price (different cost-per-placement), because there are different numbers of students in different places of the labyrinth, and the prices of problem placement would follow the supply-and-demand mechanics.

Now, to solve the big problems like aging or death, companies and research groups focused on those problems could leverage upon the resource. They could buy the placements from something like Google AdWords in the very places, where the likelihood of presence of a gamer with the right skills is the greatest.

The rest is in the details. The paradigm would work well not just because games motivate well, but also, because a part of the brain called hippocampus, involved in learning to navigate complex environments had been shown to be very neuroplastic, even for adults (goo.gl/t72Xk). (London taxi drivers were know to have large hippocampuses, but what's more impressive, new trainees also grow their hippocampuses significantly in just two years. We do have this magical piece in brain that learns as good as a young child.)

I originally posted on Halfbakery, here, and later shared on Google Plus here.