The Dawn of Collective Intelligence

Artificial Intelligence (AI) continues to dazzle us, but we mustn't forget we are also at the dawn of a Collective Intelligence (CI) revolution.

Elsewhere, I have suggested that the distributed ledger revolution includes any technologies that flow from the original ideas set out in the 2009 Satoshi Nakamoto whitepaper. Since then, a thousand competitors have vied to replace Bitcoin just in the industry Bitcoin created. The point here is not to signal any allegiances. A distributed ledger is not just a static moment but a new and evolving way to organize people that flows forward into the future. So, for our purposes, “distributed ledger” might include any number of subsequent technologies that improve upon or supplant the original blockchains. The industry is an evolving ecosystem.

One of the major improvements to the distributed ledger is, most assuredly, governance. That is, how does the technology evolve internally? Further, does it disintermediate, hypermediate, or some of both? Does it enable collective intelligence at the network level or the level of its nodes, that is, the people who use it?

With disintermediation, you eliminate middlemen. The logic of the smart contract makes certain middlemen unnecessary. But this might not be the summum bonum of all decentralization, in all cases and at all times. As holochain developer Arthur Brock suggests, the Bitcoin blockchain (brilliant though it may be) is non-social in a certain vital respect:

In computer science, an ontology describes what EXISTS in a system. For example, In Bitcoin what exists are transactions organized into blocks linked in a chain. The first transaction in each block gets to create new coins (cryptographic tokens). The other transactions spend a coin by signing (with a private key) the previous transaction to a new owner (using their public key as their address/identity). There are also nodes with which you send and receive transactions.

Notice no people in that ontology. They don’t exist.

With no people, there are no relationships, no communication, no interconnection, no community. How can a community that doesn’t exist regulate itself?

Brock points out weaknesses in certain blockchain applications and issues of governance and change. In so doing, he suggests that self-organizing systems sometimes need agents who exercise their agency. Sometimes, we want technology to be radically inclusive or to make everyone a potential middleman.

Call this hypermediation.

To sketch hypermediation, imagine that technology enables a system of numerous checkers. Those checkers use their minds to do the checking on the activities of others, but the checkers are checked in a kind of fractal. Each checker builds and guards a reputation to be considered and rewarded for future work as a checker. Such is not exactly disintermediation but rather virtuous recursion with associated good incentives. There is no perfect human fractal, but it might still be virtuous when compared to the status quo ante.

Such is not to argue that the original blockchain, with its logic fixed at the network level, isn’t profoundly useful for a host of use cases. Rather, there will be circumstances in which people will want to self-organize in ways that consider their particular strengths, aptitudes, perspectives, and context. As I write, blockchain-inspired alternatives are being developed along these lines. There will be distributed ledgers for identity, reputation, and improved trust networks.

Sometimes, hypermediation will be required. 

As these human fractals come online, they are likely to make decentralization even more potent, not less. 

Collective Intelligence

The point of the above is not to get into the weeds of the distributed ledger and its children. It is rather to hint at a bright, bold future for collective intelligence and collective action. 

Assume human collective intelligence (CI) will compete with artificial intelligence (AI). In such cases, we’d better try everything we can to see how we can build out systems of programmable incentives and radical collaboration. As we have suggested, we will eventually replace politics as the primary mechanism of societal change. This might not occur in a single moment. Rather, it’s likely to follow due to improved self-organization made possible by peer technologies.

But what is collective intelligence?

Each person has a brain, and each brain, no matter how smart, has cognitive limits—strengths, weaknesses, and unique attributes. But there are ways to harness the various strengths of various brains such that the combined effect kicks more ass.

According to science writer Matt Ridley, the market is one of the first and best examples of collective intelligence. That is, “human achievements are always and everywhere collective.”

Every object and service you use is the product of different minds working together to invent or manage something that is way beyond the capacity of any individual mind. This is why central planning does not work. Ten million people eat lunch in London most days; how the heck they get what they want and when and where, given that a lot of them decide at the last minute, is baffling. Were there a London lunch commissioner to organise it, he would fail badly. Individual decisions integrated by price signals work, and work very well indeed.

And prices—information wrapped in incentives—will continue to function as a primary mechanism of collective intelligence. 

Still, novel means of collaboration and programmable incentives mean that the market mechanism gets married to the technologies of connection. “Permit these creative know-hows freely to flow,” implores Leonard Reed’s pencil.

Busy as ants, developers are building flexible flow architectures around the clock. 

The Limits of Collective Intelligence

We should be clear about what we mean when we say collective intelligence, at least for now. CI is not some magical superbrain. While we want to leave open the possibility that, somehow, a greater thinking noosphere can emerge from connected individuals, we have to acknowledge our current limits. 

"It is a form of universally distributed intelligence, constantly enhanced, coordinated in real-time, and resulting in the effective mobilization of skills,” says cultural theorist Pierre Lévy.

I'll add the following indispensable characteristic to this definition: The basis and goal of collective intelligence is mutual recognition and enrichment of individuals rather than the cult of fetishized or hypostatized communities.

In other words, Lévy is suspicious of any computerized Gaia hypothesis or magical superbrain. And we should be, too, although we should be open to new layers of the adjacent possible, as such complexity could build another layer atop the noosphere.

Collective intelligence should also not be confused with group intelligence, though it is a subset of CI that involves small groups of people interacting effectively through brainstorming activities. Group intelligence arises in the action of, well, groups.

“We need to stop looking for leaders and start looking for teammates,” admonishes complexity scientist Yaneer Bar-Yam. “We need to find others we can trust about ideas, advice, and joint action.” That means turning away from both clumsy nation-state dirigisme and hermit-crab individualism.

Society has to coalesce into local and global teams. Teams of individuals, teams of teams, and teams of teams of teams, up to society as a whole. Whether implicitly or explicitly, everybody needs to ask: Do you want to be a member of my team? Can I be a member of your team? Can we say “we” about ourselves to become a collective, with a collective identity?

Team membership requires different kinds of decision-making and information processing. However, a team-within-team approach gets us closer to the kinds of CI humans are capable of.

One key to improved CI lies in our ability to hustle ourselves into effective teams. Another lies in the speed with which teams can acquire and use information and make decisions, as well as how we can further divide labor without being limited by geography or having our creative energies constrained by rigid systems.

Finally, improved CI will depend on expanding the marketplace of ideas to include more people while meeting them where they are. There is no magic in any of this. Instead, it means building new layers of the adjacent possible through continuous improvements to the technologies and ourselves.