The collective wisdom of the crowd is being called upon to help scientists decipher the language of pilot and killer whales in a project that could help us operate our machines in harmony with the ocean giants.

To participate, log on to Whale.fm, a project sponsored by Zooniverse and Scientific American, and try to match up similar sounding whale calls. 

---

The researchers behind the project hope that the wisdom of the crowd will more accurately match the calls than one user can alone.

That is, what I think are two similar sounding calls might not be what you think. If hundreds or thousands of people group the same sounds together, though, they're more likely a match.

Whale.fm is particularly important for basic pilot whale research, since scientists know so little about them. What they do know suggests they communicate in similar ways to killer whales, which are known to live in family groups and communicate among themselves in unique dialects.

Scientists collected the whale calls over the years with underwater microphones suctioned onto whales, dragged behind ships and attached to buoys. Each sound and a corresponding spectrogram — a visual representation of the whale sound snippet — are presented for users to find a match.

The site also plots on a world map where the call was recorded and even offers users an option to track specific whales. 

If all goes well and lots of people participate, researchers should get answers to questions such as the size of the pilot whale call repertoire, any differences between the repertoires of long and short finned pilot whales, and how, if at all, the calls change amid noise such as sonar.

The project will also let researchers know how well volunteers agree with each other and, thus, how good we are at collectively categorizing the calls of vocal species such as whales.

This type of crowdsourcing — tapping the wisdom of the crowd to form a collective intelligence — has been used for other research projects in the past, including Zooniverse's Planet Hunters, which is harnessing crowds to find new planets.

If the crowd turns out to be wise enough to help researchers decipher whale songs, what else can we do?

More on crowdsourcing projects:

• This is the first crowdsourced military vehicle
• Abort or give birth? Couple asks Internet to vote
• Software taps human brains
• Charities start to harness the power of the many
• Amazon pushes user-driven research service

---

John Roach is a contributing writer for msnbc.com. To learn more about him, check out his website. For more of our Future of Technology series, watch the featured video below.

 

Computers may eventually outsmart human intelligence, but for now they're just finally getting smart enough to ask humans for help.

That's the basic idea behind MobileWorks, a startup that is weaving crowdsourcing capability into computer software. Crowdsourcing is the concept of putting out a question to your social network to help solve a problem.

---

In MobileWorks case, software sends tasks to a hand-picked crowd — mostly workers recruited from the developing world such as the slums of India and Pakistan. Many work with a mobile phone. The company says these workers are getting high-tech experience and a "fair wage."

"Much of the criticism that has been leveled at online digital work is that it becomes kind of sweatshop labor," Anand Kulkarni, a cofounder and CEO of MobileWorks, told me today. "Our goal was to start with a livable wage and work forward to construct an effective crowdsourcing system."

And what's that wage? Workers in India on a mobile phone earn about U.S. $0.50 per hour; those with a laptop computer make $1.50.

"These are workers who are earning about $2 per day before joining our systems, so, in a way, what we are paying is enough to make a strong positive impact on their lives," Kulkarni said.

Tasks these workers accomplish include transcribing audio recordings, digitizing handwritten notes and scouring the Internet for contact information of potential job recruits. Many take just a minute or two to complete, which is part of the plan.

The cost to the user of the system is on the order of pennies per task.

To maintain client confidentiality, each task is broken up into tiny bits and distributed to the workforce. When the bits of work are completed, the software stitches them back together and delivers the completed task to the user.

The concept is similar to Amazon Mechanical Turk, where tasks are solved by a crowd of anonymous workers, though MobileWorks says their hand-picked crowd is faster and more accurate.

And since the workers are handpicked, MobileWorks can rouse them with a quick text message, making sure workers are at the ready when there is work to be done.

"The ability to spin up workers when you need them is very powerful," Michael Bernstein, who researchers crowdsourcing at MIT's Computer Science and Artificial Intelligence Laboratory who has developed an application to tap into the Mechanical Turk service, told Technology Review.

"On Mechanical Turk your tasks can just stall because not enough people chose to work on them."

More stories on crowdsourcing work:

• Facebook asks users to translate for free
• Charities start to harness the power of the many
• Cash in: 12 ways to earn
• Amazon pushes user-driven research service

---

John Roach is a contributing writer for msnbc.com.