When thinking about cloud computing, it’s easy to get caught up in the technological innovations. Processors, servers, virtualization, provisioning engines and other components continue to evolve and make computing via the cloud more efficient and more powerful. Often overlooked, however, is the tool that made all this high technology possible — the human brain. What if we could access that resource via the cloud, too?
As it turns out, we can. Some call it “labor as a service” (I like LaaS), others call it “labor-on-demand,” but everyone should call it cloud computing. Via GUI, submission form and/or API, customers designate how many workers they need, for what task and for how long, and pay accordingly. The LaaS provider handles everything else — finding, training, evaluating and paying the workers, then delivering the results to the customer.
Perhaps invented by Amazon Web Services in the form of it Mechanical Turk offering, the LaaS market is now taking off. Startup CloudCrowd this week announced $5.1 million in Series B funding for its service, while fellow startup CrowdFlower was the subject of an in-depth interview with O’Reilly Media. The spectrum of use cases for these solutions is broad, ranging from book reviews to editing to data cleansing, and beyond. If something requires human judgment that can’t be replicated in an algorithm, it’s a prime candidate for LaaS. Some providers, however, take a more targeted approach. uTest, for example, is dedicated to software testing; its testers must have software backgrounds in order to join the ranks.
In his aforementioned interview, CrowdFlower CEO Lukas Biewald draws comparisons between LaaS and cloud computing, and he’s absolutely correct in doing so (not only because I’ve drawn the same analogy in the past). Cloud computing is more about a set of capabilities than it is about any specific technology set, so who’s to say silicon has to do the actual computing? Developers turn to traditional infrastructure as a service because they can get CPU resources when they need them, for as long as they need them, and without having to go through the expense and bureaucracy of purchasing, installing and managing physical resources. LaaS does the exact same thing with employees (only application development is far easier — “Do this.”), even to the extent of letting customers choose labor by geographical location and/or qualifications.
Going forward, we should expect to see this connection take shape technologically, too. We’ve covered the increasingly tight relationships between IaaS, PaaS and SaaS offerings in offerings like Windows Azure; CrowdFlower seems intent to throw LaaS into this mix, as well. Via its RESTful API, developers can program applications to generate tasks, order work and interact with the CrowdFlower platform as data is worked upon. Imagine using a SaaS application hosted on an IaaS or PaaS platform, which automatically scales up not only computing resources, but also human resources, and works with all the data generated by every source. We’re not there yet, but it certainly appears possible as cloud resources continue to advance in terms of interconnectivity and automation.
Ultimately, I think models like LaaS will force us to look at cloud computing far beyond the current scope of IaaS, PaaS and SaaS. We’re seeing services take hold now that don’t fit nicely into the definitions we’ve ascribed to these terms, but certainly fall under the cloud computing umbrella. I’d love to hear what you think about this topic: What other types of emerging services warrant their own “aaS” acronym, and how far can we expand the definition of cloud computing before it breaks?