An Inconvenient Truth
Tales of Two Worlds · What does a decentralized world look like? (part 4)
This is part 4 of Tales of Two Worlds. Here is part 1, part 2, part 3.
—
Tales of Two Worlds
Stories of Two Worlds — through the lens of the Centralized vs. Decentralized dualism:
Centralized vs Decentralized
- Society vs. Network
- People vs. Sentient & Autonomous Agents
- Social Activity vs. Network Effect
- Clustering increases efficiency & potency vs. Clustering decreases efficiency & potency
- Organizations trump people vs. Agents trump clusters
Fourth story: An Inconvenient Truth
Cluster increases efficiency & potency
vs.
Cluster decreases efficiency & potency
Our Blind Men’s quest for understanding the hyper-dimensional Elephant, ‘what a decentralized world looks like’, has hit some great strides: the web and big data hegemony, coevolution with A.I., a new value view; these are all the projections of a hyper-dimensional world (a decentralized one) into a world we are dwelling right now — a world controlled by a Centralized Technological Industrial Regime (CTIR), all through the lens of centralized vs. decentralized dualism. The dualism angle is necessary not for the sake of storytelling, but for overcoming the mind-bending pull of the gravitational well of millennia of centralized ways of thinking, which is so strong that I fear the light of decentralized enlightenment shall never escape the crashing tow of centralized assimilation.
Ever since the start of our quest to search for the elusive semblance of a decentralized world, many are confounded by our approach — why jumping all the hoops trying to see where there is nothing to see probably? I am not surprised nor disheartened at the apparent continuous vacillation toward the enlightenment of centralized vs. decentralized dualism — considering the vast space that a hyper-dimensional world spans across, my efforts of developing just a couple of silver plates of early photos of remote galaxies are inevitably inadequate, obviously. There are so many slices of space I couldn’t just place a silver plate at it will always be an uphill endeavor to build you a ladder to reach for that image of decentralized world projection into our own world. Nevertheless I shall continue trying.
It is time I show you some specimens of true decentralized systems now. They are extremely rare. Many so called “decentralized systems” are actually not truly decentralized (and I will showcase examples as well). Then an important question would be: how do we judge if a system is truly decentralized? The answer is quite straightforward once we dissect our decentralized specimens.
Bittorrent.
Bittorrent is a decentralized file-sharing network. Traditional file sharing scheme serves requests for a file using what is called a client-server model: clients who request the file all download from a central server. Client-server model is also the linchpin technology framework for our centralized Internet: Web and Big data. On the other hand, the bittorrent protocol utilizes a peer-to-peer model: every node (called a peer) in the network downloads available pieces of a file from many nodes (peers) and reassembles the pieces back into a complete copy of the file while also uploads their copies of the file pieces to other nodes; apparently in this model there is no division of roles between a single (file) server and its clients: every peer is both a server and a client during the peer-to-peer file exchange process. Bittorrent is a full-blown genuinely decentralized system.
So what is the fundamental distinction between a peer-to-peer network and a client-server one?
First it is about the efficiency mechanism. For a centralized client-server architecture, a server’s file sharing capacity is determined as well as restrained by the combined capacity of the server itself. Remember network effect? In a client-server network it is actually called negative network effect — because each client’s effective downloading bandwidth (the prime user experience) will decrease when more clients join the network to download the file. So as the network grows in size, the quality of service degrades, and efficiency decreases as well. On the other hand, in a peer-to-peer network like bittorrent, the situation is reversed. When more peers join the network, they start to download pieces of the file from each other, so each new peer that joins the network increases the file serving capacity, therefore as the network grows in size, the quality of the service improves, and efficiency increases too. It is exhibiting positive network effect.
Second, from an engineering point of view, bittorrent, or peer-to-peer network is superior in terms of efficiency, and performance, to a client-server network. A peer-to-peer network is also more robust than a client-server network, since the latter’s network traffic flow pattern is asymmetric thus is more fragile and prone to service disruption and outage. Bittorrent is also more inline with the original spirit of the Internet — as ARPANET, the predecessor to the original Internet, was devised to be inherently robust, and could withstand massive attacks on the network, by eliminating critical choke points or paths due to asymmetric clustering of information or services, exactly the kind of weakness reintroduced to and dominated the Internet in the 1990s ever since by the business and commercial interests. As the existential threat of a nuclear war faded away and the Internet reinvented as the Web, the proliferation of the so called Internet business and e-commerce took off; it appeared that the less robust but more business friendly client-server architecture gradually took hold in the 1990s and became the canonical form of the network architecture of the Web, as well as gave way to the emergence of Big Data, the “evil” twin of the Web.
Then came the burst of the first Internet bubble of 2000. The Web in its infancy, almost got killed. Web business market crashed due to a simple technical reason — the speed and bandwidth of the Internet at the time was slow and choking; the infrastructure of the Web, which was based on the client-server architecture, wasn’t ready. And Big Data wasn’t even a thing back then. It took almost another decade for every piece of the Web and Big Data hegemony to be ready: superbly efficient server-side technologies such as MapReduce at a gargantuan scale, deployed on giant datacenter assembled on cheap hardware, and consolidated rich content ready web browsers on the client side with ubiquitous broadband Internet services. But before all that happened, bittorrent emerged shortly after the crash. It appeared to be a savior for the recovering internet licking its wounds. It improved download speed and bandwidth over slow dial-up networks almost like magic, before the proliferation of broadband services. Many aspiring Internet ventures sprinted to build viable online businesses on this nascent and innovative technology. There was also a revival of online communities back to the decentralized root of the original internet. Bittorrent almost created another path for the evolution of the Internet, one that is completely opposite to the Web and Big Data.
But the power of the One Ring could not be undone.
Unscathed from the 2000 Internet crash, a then small Silicon Valley tech startup called Google stumbled upon the One Ring, a parasitic capillary-pipetting Ad distribution system for the Web, which perfectly matched Google’s own master killer algorithm, Page Ranking, to reign in the unkempt wilderness of the Web. With the power of the One Ring, the dominion of the Web was finally established. Bittorrent’s alternate path for a decentralized Internet, was abandoned. We have lost the Internet.
How did the superior bittorrent lose to the inferior web?
Decentralized bittorrent never had the chance. The centralized web prevailed because it was meant to be. As we had discussed in previous articles, our current social-economic regime is a centralized, resource-bound one (Centralized Technological Industrial Regime, CTIR); when our centralized business world first took a look at the early Internet, they were befuddled by this plush but unruly Garden of Eden, a diverse yet messy Amazon rainforest. The problem with the rainforest is that, it is very difficult to do industrial value extraction. And the alternate value of mutual interactions within the ecosystems of the online communities of the early Internet? Totally unseen by the centralized business world, a.k.a. CTIR — because it is a completely different value system, and can only be seen if you are willing to accept a new value view, the mind’s communal energy value view. Finally came the bulldozers and the long process of deforestation — turning the rainforest into agriculture development. Thus begin the slash and burn of the early Internet, and centralized industrial value extraction of the Web. Decentralized bittorrent network never had a chance because its value view is not compatible with CTIR and can’t be seen by CTIR.
All is not lost though. Within the decade of the Web’s triumph came an alien invasion.
Bitcoin.
Hope is nigh again to build a decentralized world and restore Internet’s old glory. Looks like It is time to reactivate the insurgence. But first we must thoroughly understand the characteristics of decentralized systems and what really distinguishes them from centralized systems, in order to build a brand new decentralized world. I’ll fire away.
Bitcoin is another specimen for our decentralized system dissection. We have done some dissections before. Since decentralized system specimens available right now are extremely few and rare, let’s analyze the two decentralized systems at hand, let’s continue to dissect bittorrent, then we move on to bitcoin.
Earlier we thoroughly analyzed two different network system architectures, the peer-to-peer model of bittorrent vs. client-server model of the web, we concluded that from the tech perspective of efficiency, be it bandwidth or robustness, the decentralized bittorrent beats the centralized web, by a large margin. Moreover, from a social-economic perspective, we also observed that the decentralized bittorrent aligns with an egalitarian, free and open Internet, while the centralized web evolves into a monopolistic, controlled and walled-up “Internet”. These observations, of course, exemplifies the shear irreconcilable schism reflected through our main thesis: centralized vs. decentralized dualism. The realization of this schism hints us at another level of understanding: there is no common ground between the two extremities in the realm of this dualism. There is no common frame of reference we can use to objectively evaluate systems at the two extremities within the confinement of our world. What do I mean by that?
Now let’s revert to the centralized world view for efficiency — there you will find a completely opposite account: that centralization, or clustering, increases efficiency. In this narrative, you will find a different vector for measuring efficiency: deployment. Not that user experience (such as bandwidth and robustness) is not important, just that it is a secondary optimization target to deployment. As we noted earlier, In a centralized system, under CTIR, its existential purpose is to extract resource-bound value. In non-information technology sectors, it is the combine harvester model — the goal is to achieve the least overhead for most efficient resource-bound value extraction; clustering, like the combine harvester industrial agriculture model, increases this “deployment” efficiency by eliminating massive overhead. In the information technology sector, like the Web and Big Data, it is the parasitic model — like in biology, web users are the hosts and web service providers are the parasites in a parasitism relationship; it is about how to deploy your micro or nano capillary-pipetting Ad value extractor with the least overhead, regardless of the harm done to the host. That’s the inconvenient hard truth behind every web business model. And user experience? That is just the bait — and optimizing the bait is always a secondary concern. Before the Web grew into today’s bone-chilling duopolistic or oligopolistic hegemony, many web businesses did start as online communities, like back in the times of early Internet; but since pure online communities can’t really survive under the centralized web economic regime, these online communities all must transform themselves into the parasitic model — successful ones developed their online communities and their parasitic value extraction apparatus into a symbiotic relationship, keeping the hardship or pain of the community to a manageable level. But these symbiotic web businesses are facing extinction since the rise of the consolidated Big Data amplified Web duopolies or oligopolies. The age of the Symbiotic Web (Web 2.0 anyone?) has faded away. The current web can be best described as the Reign of Cordyceps — harming or killing the host is not just inevitable, but necessary, for the parasites’ survival. And Cordyceps’ survival depends on the extremely efficient deployment of clustering into the host body.
Phew. (cue the crescendo)
Now we can finally make an attempt to answer this quintessential question: how to distinguish a true decentralized system from a centralized one?
Armed with the insights we gained earlier, this question can be answered in two aspects:
First, from a technology or system design perspective, the primary optimization target must be some system vector that can only be optimized by increased decentralization; and a natural result of the opposite is also true: that increased centralization or clustering of the primary system vector will only reduce the optimization or worse, weaken that system vector.
Second, from a social-economic perspective, a truly decentralized system promotes the benefits of the community while a centralized system promotes the external resource-bound value extraction from the community. Furthermore, a truly decentralized system facilitates the flow of a new kind of value: a value based on the mind’s communal energy. We had discussed this earlier — the schism between the two extremities of centralized vs. decentralized dualism is irreconcilable. A “compromised”, or “hybrid” “decentralized” system could not exist — a “tainted” “decentralized” system is not a true decentralized system, it is a centralized system, there is simply no middle ground.
Now let’s move on to bitcoin.
Bitcoin is quite different from bittorrent. Bittorrent matches our description of a pure decentralized system perfectly: its optimization goal — the system vector comprised of bandwidth as well as robustness, increases as more users join the network to access such system vectors. That’s the technology and engineering aspect. Social-economically, bittorrent networks promote community benefits — user experience, free, open and egalitarian community participation — not external resource-bound value extraction at the cost of the community. Bitcoin core protocol and network also fit all above. but Bitcoin is more complicated than bittorrent.
You see, bittorrent is a simple, self-contained decentralized system. The communities built on bittorrent is also self-contained and self-sufficient. Bittorrent fulfills its network participants’ need and no outside forces need it depend on. As we discussed earlier, the whole Internet could have been built on bittorrent without any external forces, as far as we know — there do still exist fringe online communities built solely on bittorrent and are self-sufficient. But that is not the case for bitcoin.
Bitcoin is a pure decentralized system, but it is not self-contained. Since its early inception bitcoin has fused with very centralized forces, such as bitcoin exchanges. Many bitcoin network participants, such as miners, mining equipment builders, software developers, are also increasingly exposed to centralization influences. In fact, although the core bitcoin network is still decentralized, the close-knit bitcoin ecosystems, as well as all cryptocurrencies and their ecosystems inspired by bitcoin, are characteristically centralized.
This is a sad situation for bitcoin. But it is not bitcoin’s fault. The cause for this situation is also two-fold. First, bitcoin is not self-contained because it is only a projection from a hyper-dimensional decentralized world. The core features of bitcoin all hint at a certain receiving end or infusion joint of a decentralized nature. Unfortunately, we can’t find corresponding decentralized systems to interface or infuse with bitcoin. This brings us to our second point: our world is centralized and our world consumed and wrapped up bitcoin with layers upon layers of centralized slime, making a Frankenstein monster out of bitcoin that is very centralized.
The different behaviors of bittorrent and bitcoin when coming in contact with our world, which, come to think of it, could also be fantasized from a chemical perspective. Both bittorrent and bitcoin are very simple decentralized systems. Let’s imagine them as metals. Metals can display many characteristics with direct assemblage of decentralized networks: they are malleable and ductile, deforming under stress without cleaving (sounds like the desirable properties taken from a page of the original ARPANET design document, does it not?); these characteristics can be attributed to the isotropic, or nondirectional nature of metallic bonding, the prominent chemical bonding that gives metal its chemical and material properties. You should know that the aforementioned characteristics are not exhibited by most materials people encounter in nature, just like people haven’t encountered true decentralized systems until quite recently. Although metal in general is not rare in the earth’s crust (about 25% abundance), metal in free elemental (native) form is relatively rare — such as gold. Gold is the most noble of the noble metal, resistant to corrosion and oxidation — we can compare this to bittorrent’s self-containment characteristic; now just imagine: an internet made of gold, how does that sound?
Let’s continue the chemical analogy for the self-containment probe. If bittorrent can be compared to a noble metal, then bitcoin can be compared to a base metal, which is easily oxidized or corroded — just like crypto exchanges corrupt bitcoin’s decentralized nature.
Am I torturing you mind with this chemical contortion? Maybe. I believe the bittorrent and bitcoin metal metaphor sketches an origin story for decentralized systems: that they are very rare but potent, they interact with the world they come in contact, and their emergence will change history profoundly — just like metal completely changed the face of human civilization, and put us on a steroidal path with no return. A decentralized world is the future for humanity.
This concludes our dissection of bittorrent and bitcoin, our only specimens of true decentralized systems. What now? We have literally come to the edge of knowledge.
It is time for us to cross over to the realm of the unknown.
Feeling a bit like Captain Jack Sparrow on Black Pearl when it sails toward World’s End? We know our centralized world is racing to a cliff — that the duopolies and oligopolies of the world shall soon harness the power of AI to enslave humankind. You may think that fictions are pointless but I digress. Nevertheless, standing at the beginning of history is both exciting and disheartening.
Human civilization has come a long way building a cathedral of a centralized world. That world is both towering and terrifying at the same time. And the decentralized world? Bittorrent and bitcoin, two peanuts (among other peanuts). So insignificant. So inconsequential. You tricked me into a luscious dream. I was promised the Garden of Eden. And a hyper-dimensional one at that. Now I only see two peanuts (like Captain Jack Sparrow in one of his demented dreams). I want to wake up and return to the real world.
No, you don’t. The centralized “real” world is the dream. The decentralized world is much older, larger, and real. The centralized “real” world humans built? A cathedral built on paper. Insignificant, inconsequential — if you know how big and fascinating the decentralized world is, old and new.
Let’s imagine a world full of decentralized systems with much more complexity than bittorrent and bitcoin. And not just 2, but close to 9 million different decentralized systems, and billions of times more complex than bittorrent and bitcoin. And the number of “nodes” in all of those decentralized systems? How about around 1 trillion?
Yes, you should’ve already guessed it. Let’s leap from inorganic simple metal, to life. Life forms, if you count from the smallest unit of replication, the cell, then the total number of cells in all species on earth would be astronomical, many magnitudes bigger than 1 trillion. But let’s not get bogged down by numbers. Biological systems are strictly decentralized. No matter what weird or marvelous wonders of life forms you pick, they will always function by installing node software (DNA) onto sufficiently large number of network nodes. If we check the mechanisms of life forms, they will match our earlier criteria of being a decentralized system perfectly, both from engineering and social-economical (or let’s say communal) perspectives. What’s more, how about the jewel in the universe, the human brain (or any brain, not just human’s)? It’s hard to imagine it is not decentralized — there are, after all, 85 billion nodes (neurons) screaming to tell you that it is decentralized — or if it is not — then where are the Google and Facebook, duopolies and oligopolies inside the brain to pull the strings and bend your free will like you are a zombie? I guess nowhere. Which comes down to this: your very existence is decentralized. Otherwise you will not have free will. Free will is decentralization.
Design is an illusion. Centralization is an illusion. Wake up and get to work. We have a new world to build. A decentralized world.
