Society is obsessed with their health and how they look. With people paying loads of money for experts for nutritional and dietary counsel, more and more are looking for an easier way to watch how they eat.

That is why computer scientists at the Harvard School of Engineering and Applied Sciences have developed a tool that revolutionizes how a person takes eating advice - through a crowd.

The tool, called PlateMate, allows people to take a picture of what they are eating and send it to group of impartial observers without having to pay a dime for the advice.

The developers of PlateMate coordinated with the Amazon Mechanical Turk, a system originally intended to help improve product listings on Amazon.com. In this system, a group of people, called Turkers, receive a few cents for each puzzle-like task they complete.

It is with the Turkers do the people using PlateMate send their picture to for estimation of nutritional value. The group of people then tries to distinguish between foods in the photo, identify what they are, and estimate quantities. The nutrition totals for the meal are then automatically calculated.

But this system is not without problems. For instance, Turkers committed errors in identifying the foods in the photo taken, thus giving faulty estimated quantities. The contents of the foods as well as its size were also subject for error since the group of viewers did not know how to tell them with just the picture itself.

Having encountered these problems, PlateMate's developers Jon Noronha and Eric Hysen solved them by designing simple, clearly defined tasks and algorithms that compare several answers which then selected the best one. This way, errors and human faults were avoided.

However, the developers pointed out that it may not solve the challenge of eating well, and stressed that their attempt is one of the first steps to use multiple human-computational approaches to solve complex problems.