In Michael Nielsen's book, Reinventing Discovery, he performed an extensive survey of attempts to amplify collective intelligence and then presented four principles that any group process should possess. BrainSwarming meets all these requirements; brainstorming does not satisfy any of them.

(1) A rich and well structured set of information common to the group that people can easily build upon.

The BrainSwarming graph presents the commonly available information that people can easily build upon. With brainstorming, you need to remember what people say and a textual summary is sent around after the session is over--which is too late.

(2) Microcontributions (i.e., small contributions).

Contributions should be brief so they can be read quickly and a new person can easily "get up to speed" on the status of the problem. BrainSwarming uses Post-Its upon which only a small amount of text can fit--much as Twitter's 140 character limit forces brevity and encourages self-editing. With brainstorming, people can talk for long periods of time unless controlled by a facilitator. The silence and written Post-Its of BrainSwarming keep the talkative people under control with no need for a facilitator.

(3) High Degree of Modularity (i.e., break problem into very small parts)

Contributions should represent very small parts of the problem. The BrainSwarming graph divides problem solving into its natural joints of small parts of the problem. A contribution is either a goal, sub-goal, a resource, a part of a resource, a feature of a resource, or an effect achieved by interacting resources. With brainstorming, there is no structure to a person's contribution so people tend to share complete solutions or nearly-complete solutions. However, a complete solution hides many alternative solutions that should be articulated.

For example, in the Harvard Business Review video on BrainSwarming, someone might want to solve the problem "remove ice from power lines" by "shake the power lines with a long pole." 

http://blogs.hbr.org/2014/03/why-you-should-stop-brainstorming/

BrainSwarming would divide this solution into small components (i.e., "shake the power lines," "long pole," and a line connecting the two indicating that the long pole can be used to shake the lines). These small components on the graph make it clear that many possible alternative solutions exist. For example, entering "shake the power lines" as a sub-goal makes it clear that there are many ways to induce shaking: strong controlled wind from a helicopter, blast of sound from a sonic gun, a small robot that crawls across the lines, etc.

(4) An architecture of attention (i.e., get people with the right expertise to the right part of the problem). 

In the online version of Brainswarming, users register their areas of interest and expertise. The software guides users to parts of the graph that match each individual's interest and expertise. With brainstorming (even electronic brainstorming), this requirement is usually not considered.

Mix together insights from insect problem solving, human social and cognitive psychology, and machine learning and AI (artificial intelligence) and you have a potent new model for problem solving and innovation called online BrainSwarming.

Harvard Business Review's video on BrainSwarming has caused a minor stir and I have been swamped with requests to train people on the finer points of conducting this new group process that is modeled on how insects solve problems together (i.e., Swarm Intelligence).

http://blogs.hbr.org/2014/03/why-you-should-stop-brainstorming/

By eliminating talking and having participants contribute to a structured graph visible to all, BrainSwarming has been shown to produce more ideas than brainstorming and in less time. Just as insects leave signals in their environment that influence other members of the swarm, BrainSwarming participants leave briefly-stated ideas on a structured graph for others to build upon. Goal refinement grows downward (see the Figure) and resource interaction (and resource addition) grows upward until the two directions meet, which indicates that solutions are starting to emerge (i.e., people are using the resources to accomplish the goal).

Since people are working in parallel (and not sharing one at a time), the work proceeds much faster than in brainstorming. Since people work in silence, outspoken people cannot dominate the process and curb idea generation. Further, no facilitator is required to stop people from dominating and critiquing--the silence does this implicitly. There is no need to have a scribe record everything and send it around later (usually as a long list of text)--the group works together to record and organize everything on the graph. The graph can stay up on the wall so people do not have to be physically in the room at the same time to work on the problem.

BrainSwarming's silence is ideal for introverts (nearly half of the population): http://ahauniverse.weebly.com/1/post/2014/04/introverts-shine-in-the-silence-of-brainswarming.html

BrainSwarming actually fulfills the original goals of Osborn much better than brainstorming by keeping the good of brainstorming and getting rid of the bad:
http://ahauniverse.weebly.com/1/post/2014/04/brainswarming-keeps-the-good-of-brainstorming-and-gets-rid-of-the-bad.html

Online BrainSwarming is ideal for remote groups working online. Imagine a manager posting a new graph one night with merely the goal at the top and the known resources at the bottom. Imagine this manager waking up the next morning to see what contributions came in from colleagues in Paris, Australia, Singapore, etc. The graph keeps track of all contributions, how they relate to each other, and whether any solutions emerged (i.e., connections between the downward growing goal refinement and upward growing resources).

Online BrainSwarming Incorporates Machine Learning Techniques

The full power of online BrainSwarming only comes into effect when software incorporates my Obscure Features techniques as well as Machine Learning techniques. Since any innovation is built upon what is commonly overlooked (i.e., obscure), all my techniques for uncovering the obscure will run in the background and reveal hidden features, assumptions, and connections as the graph is being constructed. (Contact me to receive a published paper in Psychological Science on my Obscure Features techniques: tony.mccaffreyphd@gmail.com.)

Further, Machine Learning techniques will mine the history of all problem solving activity at the site to find solutions from other problems that are "like" (i.e., analogous to) the current problem. (Contact me to receive a published paper on these particular Machine Learning techniques: tony.mccaffreyphd@gmail.com.)

An eBay or eHarmony for Problem Solving

The final result is a site where some people come looking for solutions to their problems and others come looking for problems to solve. The result is a marketplace or "dating site" where problems and solutions find each other. It is like InnoCentive only with Obscure Features and Machine Learning techniques to make it better. (Contact me to receive a published paper on a fuller vision of this platform: tony.mccaffreyphd@gmail.com.)

BrainSwarming Satisfies Nielsen's Requirements

Michael Nielsen, in his book Reiventing Discovery, surveyed the vast array of attempts to problem solve together online or as he likes to say amplify collective intelligence. Nothing that Nielsen surveyed satisfied all of his requirements, but BrainSwarming does!  (Contact me to receive a published paper on fulfilling Nielsen's requirements: tony.mccaffreyphd@gmail.com.)

Summary

In sum, the combination of insights from insect problem solving, human social and cognitive psychology, and machine learning and AI (artificial intelligence) make for a powerful new model that could transform how humans problem solve and innovate together.

Contact me to receive published papers on any of these aspects as well as the status of my fundraising efforts to make online BrainSwarming a reality: tony.mccaffreyphd@gmail.com.