Gaming skills harnessed for good

Gamers have long been derided for their skills not amounting to much, that skills learned through gaming do not have much application in the real world.

It’s not completely true, of course, because some games have had sociological side effects that are beneficial but now there’s something new.

As far as existing gaming being leveraged, teachers have long utilised specific games that engender puzzle solving (The Secret of Monkey Island comes to mind) but it’s still a passive exercise.

Now enter the concept of FoldIt, actively using gamer skills.

The page explains it in much greater detail, but I’ll sum up where I’m coming from.

One of the problems relating to trialling drugs and fighting diseases is being able to model how proteins and drugs will interact, because protein chains are devilishly complex. Not only in themselves but in terms of modelling them, especially for a computer.

But contrast that to a human brain. Human brains are massively powerful computational problem solvers – specifically in terms of finding associations and patterns in things, a skill that is demonstrated to be useful in such problems, and a human can make associations far faster than a computer can.

That’s something that computer scientists already know, of course, but most people just don’t realise it. For example, give a list of numbers to a human to put them in order, and the same list to a computer. The human can intuitively process the list and come up with a likely-correct solution very quickly (in a few passes) while multiple different methods exist for a computer to do it, and it wouldn’t know which to use without instruction, not to mention that it wouldn’t be able to make any associations about the data.

Computers are very good at going through data, looking for specific items, or applying specific rules and methods to them but they have little problem solving capability (entire branches of mathematics and computer science are dedicated to solve some of these problems that a human could approximate a likely-reasonable solution to)

That is, of course, unless you want to let the computer go through every single permutation which it will given the chance. For arbitrarily small problems, it’s workable, but for anything on the scale of real world problems like protein folding, any approach that speeds up the solving process must help.

So enter FoldIt, which approaches the protein-folding issue by turning it over to gamers. People generally play games to win, it adds a sense of competition, and harnesses all of the natural abilities people have to try solving problems.

It’s not likely that any given problem will be solved perfectly optimally at the first step but with that sort of problem, there is not necessarily an optimal solution that can be found quickly – but the more problem solvers thrown at a problem, especially where they are using their own natural skill, the chances are that more and more optimal solutions will be found over time as people find patterns that may not have been identified before.

I have to admit, while I’ve seen the power of crowd-sourcing before, of taking a single massive task and distributing it over many contributing solvers (e.g. the massive amount of data in the SETI project being sent out into different individual computers to solve) but there it’s mostly contributing arbitrary spare compute time, rather than users actively contributing their own problem solving skills, which I don’t recall seeing before, but I can only hope this is an innovative approach that we see used more and more to solve problems that can’t be solved computationally, at least not efficiently.

This entry was posted in Uncategorized. Bookmark the permalink.