Sunday, November 25, 2012

Crowd Models Become Pervasive Across Society

Crowd-based models are becoming so pervasive that almost no major segment of modern life is left untouched by them. The concept of digital crowd models refers to the coordination of large numbers of individuals (the crowd) through an open call on the Internet in the conduct of some sort of activity.

Not only are crowd models an efficient at-scale alternative to former methods that the Internet now affords, but at another level, crowd models are also a node of progress for humanity, both individually and en masse. An inherent property of crowd models is greater autonomy, decision-making, and action-taking by the individual. This means greater individual agency whilst simultaneously enabling group collaboration projects at previously unimaginable scales, for example, possibly ultimately coordinating and employing the cognitive power of millions of human agents.  

Crowd Models by Sector
  • In economics, there are crowdsourced labor marketplaces where simple tasks and professional services requests can be posted and fulfilled, crowdfunding websites for financing projects, and group purchasing mechanisms. 
  • In politics, crowd models mean the use of big data and social media to organize opinion and action, conduct direct marketing, and affect change. 
  • In the social venue, blogs, social networks, and online dating sites are examples of crowd models. 
  • In entertainment, there are massively multiplayer online games and virtual worlds. In education, vast eLearning networks are populating the landscape. 
  • In health, there are health social networks, digital health collaboration and experimentation communities, crowdscience competitions, and new movements such as the quantified self. 
  • In the legal venue, digital public goods have arisen through crowd contributions such as the Wikipedia, online health databanks, and other data commons resources.

Sunday, November 18, 2012

Supercomputing Increments Towards the Exaflop Era

The November 2012 biannual list of the world’s fastest supercomputers shows the winner incrementally improving over the last measure. The Titan (a Cray XK7, Opteron 6274 16C 2.200GHz, Cray Gemini interconnect, NVIDIA K20x) is leading with 17.6 petaflops per second of maximum processing power. This was only an 8% increase in maximum processing speed as compared to other recent increases of 30-60%, but a continued step forward in computing power.

Supercomputers are used for a variety of complicated modeling problems where simultaneous processing is helpful such as weather modeling, quantum physics, and predicting degradation in nuclear weapons arsenals.

Figure 1. World's Fastest Supercomputers. (Source)
Increasingly, supercomputing is being seen as just one category of big data computing along with cloud-based server farms running simple algorithms over large data corpuses (for example Google’s cat image recognition project), crowd-based distributed computing networks (e.g.; protein Folding@home with 5 petaflops of computing power, and crowdsourced labor networks (e.g.; Mechanical Turk, oDesk, CrowdFlower - theoretically comprising 7 billion Turing test-passing online agents).

Monday, November 12, 2012

IOT Appliances Blur the Distinction between Matter and Man

The growing wireless Internet of Things (Sensor Mania!) could bring a ‘Cambrian explosion’ in wearable computing and the number and types of Internet-connected sensors, devices, hardware platforms, software programs, and end-user applications.

There could be an adjustment period as humans adapt to an Internet of Things (IOT) landscape with more kinds of data and different mindsets, activities, behaviors, and perspectives when interacting with these data.

Whole fields of study previously limited to self-reported information such as psychology could be radically supplemented and transformed with objective metrics obtained from the IOT.

The IOT is in the early stages of modulating data onto the world of existing artifacts.

Increasingly objects may be able to collect their own data and act on it autonomously with pre-set limits and degrees-of-freedom algorithms.

Eventually, the IOT label could become a redundant demarcation as all human-manufactured matter in the future could have integrated sensors and microprocessors.

A next generation of sensors and microprocessors is already being developmentally fashioned from organic, inorganic, and hybridized material, using cutting-edge technologies for manipulating organic and inorganic matter such as synthetic biology and molecular nanoelectronics.

Distinctions between man and machine, and subject and object could blur further as IOT appliances eventually create a layer of exosenses to greatly extend current human capabilities and the ability to integrate with the outside world.

Monday, November 05, 2012

Crowdsourced Labor, Digital Marketplaces, and the Future of Self-Actualized eWork

The third annual CrowdConf on changing the future of work through Internet-based crowdsourcing labor models was held in San Francisco on October 23, 2012. The field is much larger compared to prior years (2011 and 2010), both practically and intellectually, as a range of industry vendors and other ecosystem members (e.g.; research organizations, foundations, financial community) attended the conference. Participation in crowd-based models has been increasing this year. Crowdfunding (raising money to back projects through an open call on the Internet) platforms like Kickstarter and indiegogo have been used to raise millions of dollars for projects (although both of these sites have announced that they will not be launching a new development, equity crowdfunding). Crowdsourced labor marketplaces too have been growing significantly.

What is Crowdsourced Labor? 
Crowdsourced labor is sourcing human workers via the Internet for any variety of tasks or work product, with or without remuneration, usually including an online system for bidding, quality tracking, and reputation feedback. There are digital labor marketplaces for general tasks, both short activities (e.g.; to recognize images, validate information, create electronic forms, read handwriting, classify data, tag images, transcribe audio, translate languages, and verify business listings) and professional services engagements at sites like CrowdFlower, CrowdSource, oDesk, ClickWorker, Mechanical Turk (from Amazon)) mobileworks (including real-time task-routing to the best candidate online right now), and vocational specialties such as software programming (TopCoder, Elance, vWorker (was Rent a Coder) and Guru), graphic design (99designs, crowdSPRING), writing, editing and proofreading (CloudCrowd, Soylent), document processing (microtask), and customer relationship management (LiveOps). Mobile phone-based task completion is an important category of crowdsourced labor with projects making use of time-cycles while on the go, and sites like Gigwalk ('hire your smartphone army') offering location-specific tasks.

Figure 1. Real-Time Barometer of CrowdFlower's Digital Labor Market (source: CrowdFlower)

Crowdsourced labor also includes uncompensated work in the volunteer, gift, and reputation economies in examples like Twitter’s localization service, the Wendy Schmidt Oil Cleanup X Challenge (a $1.4 m X Prize for solutions to accelerate oil spill cleanup), and crowdscience games and competitions for protein folding prediction, and genetic and other science data analysis at sites like foldit, EteRNA, Phylo, and Kaggle.

The digital marketplace concept extends more generally to that of resource allocation, allowing supply and demand to meet in real-time on the Internet in other models such as a labor markets for physical-world tasks (e.g.; TaskRabbit and Zaarly), and resource allocation (e.g.; AirBNB (apartment rental) and Uber (private driver ride service on-demand)).

Microtask Design 
Crowdsourced labor is starting to be seen as a general resource that any individual or business can tap on demand. To do so, it is helpful to know how to design a microtasking project, which can be done either independently or through third-party vendors like CrowdFlower and CrowdSource. At the conference, YP (Yellow Pages) shared best practices from a photo moderation project: they used Mechanical Turk’s Master’s Pool of photo reviewers (96% accuracy vs. 80% for unqualified task-completers), had a batch size of 40 photos (more produced worse quality), 8 buckets of keep-suppress gradations for evaluation, and a price of $0.04 per batch (1/6 the price quoted by a third-party vendor).

The Future of Work 
The conference considered more broadly the future directions of work, labor, and organizations. One analogy is that similar to the cash register being a decisive node that allowed business to explode from small family-owned, trust-based activity to large-scale enterprise, crowdsourced labor could be a similarly explosive node. Productive enterprise is no longer confined to physically-based organization-internal corporate structures. Management has been a technology, and crowdsourcing and algorithms are disrupting it. The new challenge becomes how to build and structure work for large-scale crowdsourcing, for many kinds of productive enterprise. There are new capital streams (e.g.; social, political, intellectual, and affective) available to be directed for productive construction on a global, flexible, fungible, and accountable basis. Workers and more generally humans can be put together in new ways, likely in increasingly automated ways to direct individuals to projects in areas of personal interest, for example rated on a self-actualization and contribution scale.