Network Currency

Net neutrality is on the chopping block again. Network packets will soon no longer all be created equal. Those with financial backing will be given a fast lane. Make no mistake, that financial backing will come from you. For example, Netflix, et al, will have to pay the toll to keep their service responsive and competitive, and they’ll pass that cost right on to you, their customer. The decentralization and democratization of network connectivity and services forms part of the design basis for the Personal AI we are building at Presence. Let me take you through why this is so.

In some sense, net non-neutrality is actually a reality we have been living with for quite sometime. ISPs offer connections with a bandwidth ceiling and charge a premium for higher bandwidth service. Not only that, unlike other utilities, you pay whether you use it or not. For example, if you buy a 50Mbit bandwidth service, you pay the same amount whether you are streaming HD video 24/7 or reading email once a week. In other words, bandwidth is throttled not metered, although the technology exists to do so. As an analogy, imagine that your electricity service was billed at a flat rate, no matter how much you happened to be drawing at any given time. Whether you were running your electric clothes dryer or an LED light bulb, you would pay the same amount. Now imagine that the maker of the clothes dryer was forced to pay a premium to the owner of the power grid to provide an adequate clothes-drying service to you. That’s net non-neutrality.

In response to looming net non-neutrality, discussion of mesh and peer-to-peer networks is heating up again. I first encountered mesh networks in 2011 through the work of the Free Network Foundation. The idea behind these mesh networks was that people’s devices could talk to each directly rather than relaying all interaction through centralized infrastructure and services. It also meant that a single internet gateway could be effectively shared among all the interconnected devices even if only one of them had a relationship with the provider of that gateway (the ISP).

That encounter with mesh networks, and the needs of the people using it, got me to thinking about why networks form and why they die, in other words, the application of networks. Networks form to fulfill the needs of its members, and live as long as they continue doing so. When those needs are no longer fulfilled, or if the network is unable to adequately fulfill those them, or if trust in the network erodes, then the network dies. It is no longer needed, and new networks that are up to the task will replace it. This is a natural process. It closely resembles what we see in nature and with human groups forming, flourishing, and dying off over time. I captured my thoughts on all of this in a blog and in some prototype JavaScript code over on in 2012.

The self-organizing, emergent networks I envisioned for netfoundry have some fundamental differences from mesh networks, however. Netfoundry networks are application-level peer-to-peer (P2P) networks. While mesh networks typically involve lower level protocols and connectivity structures, P2P refers to applications that connect users directly to each other rather than through an intermediating third party service regardless of how the lower level network connectivity is achieved. P2P networks are commonplace these days. Voice-over-IP and web conferencing services like Zoom use P2P techniques, as do cryptocurrency, file sharing, and chat applications. While mesh networks remove the ISP intermediary, P2P networks remove the intermediating service provider. The combination of the two creates an entirely decentralized network.

Presence is evolving the thinking behind netfoundry at the network application level. A Personal AI will have the ability to connect directly with other Personal AIs, without going through any centralized intermediary. This will result in a P2P social network of artificial intelligences. Our design for Personal AI encodes your needs as the prime directive of its algorithms and optimizations. Since it has connections to your friends’ Personal AIs, and thus by extension, to your friends-of-friends’ Personal AIs, and so on, you benefit from having an agent on a vast network whose directive is to get you what you need and shield you from distraction. Whether you need to talk to someone who has expertise in, for example, 16th century horsehair paintbrushes, or someone who can help you rid your computer of malware, your Personal AI will be searching on your behalf.

It’s not all about you, however. Getting what you need usually requires that you give back in some way. In fact, the network depends on it. A network in which no one gives back is a network that does not fulfill anyone’s needs. Such a network would quickly die. What do you give back? You give your skills, information, strengths, expertise, and interests. Your Personal AI will help you make available what others may be searching for. If you are an expert in 16th century horsehair paintbrushes, for example, your Personal AI can automatically find others interested in the same topic.

Networks interconnect those who offer something to those who need it. Thus, the relationship between a need and its provider is the unit of value, the currency, of a network. The more interconnected the network, the more likely it is that what you need, no matter how obscure, is only a few network hops away. Strong relationships create strong networks.

I hope this has provided some insight into where we are headed with Presence. The end of net neutrality will motivate new thinking about how networks are built. One of the core principles of the design of the internet is that it routes around bottlenecks. If money becomes the bottleneck, then the structure of the network will change to route around it, otherwise it will die. The strength of the network is how well it enables mutually beneficial relationships. P2P and mesh network are generally neutral, democratic, and immune to financial leverage. We are preparing to get your Personal AI connected to its peers.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>