We've come a long way in describing the work we do here in Forge Lab and more broadly at the Institute itself. What we've settled on is this: we are engineering collective intelligence networks. If you don't know how the internet works, or what collective intelligence is this statement is a little opaque but hopefully this helps:
This is a (very) basic diagram of how the internet works. On one hand, you have your input/output console, and don't let the verbiage fool you, that's the device you're using right now or anything else facilitating an internet connection. This device probably has some auxiliary equipment-- a camera, printer-- you might even have an auxiliary port to plug something in temporarily. Auxiliary devices may have data storage capability, but your device certainly does. That data might be used for a variety of things. It might contain a program that executes code to mine Bitcoin, or it might just be cat photos. Either way, your device also has a means of accessing, displaying, transforming and transmitting data.
There are billions of input/output consoles. They connect to data storage devices and each other through the internet. Some people argue all these input/output consoles collectively make up the internet, but the statement ignores the infrastructure all these devices use to communicate. According to CSU professor Larry Press, that infrastructure--wires, cables, etc.-- cost the US government $125 million, and that doesn't begin to account for the 5G and other towers operated by commercial telecommunications today. What's more, the internet doesn't just send messages. Because of mass data storage you can also "hang" a message as a social media profile/post or a website. Most of the data in messages and websites isn't stored on your device, but in the cloud-- the data storage on the right side of our diagram. For example, NTARI.org is stored in a data center in Virginia.
On the other hand you could absolutely claim that every device is the essence of the internet. Imagine all of the knowledge on the internet as a ball in space. Now imagine all of the knowledge of mankind on or off of the network. As long as you don't count the billions of copies of your favorite memes, the latter ball (all knowledge) is much larger. That is why when AI are trained on internet data they don't turn out very smart. The ARPANET was decommissioned in 1990 but what remains is the algorithm it used to move information around-- transfer control protocol/internet protocol or TCP/IP.
TCP/IP isn't "owned" by anyone. It is an open-source program anyone can use to move information around. It works a lot like the postal system, only much faster. Instead of putting your message on paper with ink...sealing it in an envelope...placing it in the mailbox...waiting for the postman...and so on...until the letter is delivered 3 days later by a different postman, your message travels at 186,000 miles per second, directly to its destination.
Most of us can't grasp what 186,000 miles per second means on intuition alone. From where I am in Louisville, Kentucky, USA, in one second I can send data to Baker Lake, Nunavut, Canada. "We've been doing this for years," you might say unimpressed, but consider the internet only came online in 1969 and full commercial access to it is a phenomenon that is not complete even today.
What does this technology mean for community, when we can talk to someone so distant as if they were so close? How do we relate in a world where relationships have always been based on the ability to communicate locally? We've explored a lot of thinking on how the internet will change society. Of note in this case is Balaji Srinivasan's The Network State. For several months this year, NTARI was promoted as a network state. It fits key definitions Srinivasan argues, but it also breaks out of his mold in other areas. We're not forming a society around a cryptocurrency, we're not purchasing land for new communities, and the founders of our organization are ultimately of little consequence. There's much more we disagree with Srinivasan on, but ultimately he does a great job of showing us a novel idea-- organized groups of distributed people.
The Network Theory Applied Research Institute is not a network state, we've realized. We're just a research institute, but the subject of our research and development will create systems that resemble or even support network states. Divya Siddarth and others at the Collective Intelligence Project call the objective of these systems collective intelligence (CI). They make strong, technical arguments against network states, opting instead to support the idea of network societies. Their definition for CI is:
Effective, decentralized, and agentic decision-making across individuals and communities to produce best-case outcomes for the collective.
We don't reinvent wheels (on purpose) at NTARI, so we use this definition to support our design philosophy. We realize what we're building are not just collections of 1s and 0s. They are methods of capturing, processing and transmitting information. Networking is conversation. We used to call the CI systems we're thinking about human-based computation networks. Thinking in this vein, we used to compare our systems to the Amazon Mechanical Turk program. Users in the program complete tasks computers can't, in some cases compiling data that will later teach a computer to recognize what it could not before. This is nearly textbook human-based computation (if Wikipedia is the textbook).
Since familiarizing ourselves with the Collective Intelligence Project's line of thinking, we've started using the term collective intelligence network. The reason actually lies in the textbook reference to microwork. Another name we threw around from time to time was mass production network. With this term we referenced the effect of gig-economy apps like Door Dash, Uber, Airbnb and more. Users are threefold-- restaurants, consumers and drivers-- working together in geographic regions to achieve a shared product or service objective. That is what we intend to facilitate with collective intelligence networks.
The Network Theory Applied Research Institute researches communication theory and information technology to create collective intelligence networks, but networks are not just digital as the Collective Intelligence Project eloquently point out:
"Decades before the construction of the first precursors to the internet, social, physical and biological scientists began to use networks to move beyond the simplistic models of discrete and infrequently interacting atoms. Quantum physics replaced the simple billiard balls of classical mechanics with complex partially entangled patterns of particles. Ecology and systems biology enriched simplistic theories of “survival of the fittest” by highlighting the network of symbiotic relationships and ecosystem services that determine the success of ever higher levels of life. Modern neuroscience, and the work in artificial intelligence that builds off it, replaced traditional accounts based on logical deduction with “connectionism”, where intelligence emerges from networks of simple but adaptable interactions of neurons."
When users all over the world can get behind an idea, they form a network. We've seen this several times in recent history-- on January 6, 2021, during the Arab Spring, the Me Too Movement, and much more. As we develop networks at NTARI we recognize the importance of respecting human rights as we manufacture the new standards of CI networking. We also recognize that networks take on many forms-- family, neighbors, nations. We endeavor not to meddle in user affairs, but deliver networks that allow them to speed vital information and processes wherever they need to go in the global community. That is what's possible with the internet, and it's about time we do it.
As collective intelligence communities emerge, employing the internet to its maximum capacity, the Network Theory Applied Research Institute will be there to support. Support our research and development by visiting
Comments