Centralization vs Decentralization

Decentralization is one of, if not the most, discussed features of the crypto tech stack. In a decentralized system, no single body controls the system. We have most certainly not reached the era fully decentralized systems, but that is what most of the world-class technologists working in the crypto sector are focused on getting us to and I believe we will get there in the not too distant future.

If you are a student of tech history, you will not be surprised that decentralization is also the right technology arriving at the right time to solve some of the most challenging policy problems facing the tech sector right now.

Before I elaborate on that, I want to show you a slide from my colleague Nick‘s deck on crypto that he uses to talk to policymakers and elected officials. I believe he borrowed it from our friends at Placeholder and they are credited at the bottom of this slide.

Here is my quick explanation of that slide.

IBM had a near monopoly on computing by virtue of their domination of the mainframe, mini-computer, and, it seemed, the PC computing platforms.

But the open PC hardware standard allowed Microsoft to develop an operating system that could run on any computer built to the PC hardware spec and they eventually unseated IBM, only to become a near monopoly themselves.

But just as we were wringing our hands about what to do about Microsoft’s monopoly, an open source operating system (Linux), the internet protocols, and the free distribution of the world wide web undid that monopoly and we got Google, Facebook, Amazon, and other big tech platforms.

And now we are wringing our hands about these near monopolies and their market power and the ability of bad actors to manipulate them. And around the same time, the technology to architect and scale a completely decentralized system emerges.

The other thing that is true of these moments of hand wringing is that just as the technology is emerging to unseat the near monopolies, regulators and elected officials try to put the genie back in the bottle using traditional regulatory techniques that often end up more deeply entrenching these near monopolies.

To give you an example of how this might happen, I am going to suggest you all go read my partner Albert’s post from yesterday on Twitter and how one might approach addressing some of the vexing problems that platform is dealing with right now.

Albert points out that:

On the minus side the calls to treat Twitter as a traditional publisher are growing.

That is how elected officials and regulators often think. They look backwards to find a model of regulation that has worked in the past and try to apply it to a new thing. But as Albert explains:

The idea that there could or should be a single central institution, let alone a commercial company, which as a benevolent dictator resolves all of these issues to everyone’s satisfaction is a complete non-starter.

Instead he proposes a few ideas that are steeped in decentralization:

my preferred go to answer is to shift more power to the network participants by requiring Twitter (and other scaled services) to have an API. That would allow endusers to programmatically create the best version of Twitter and would also make it easier to simultaneously use Twitter and new decentralized alternatives.


Twitter should significantly expand the features that let individuals and groups manage the visibility for tweets for themselves. There are already useful features such as muting a conversation or blocking an individual. These could be expanded in ways that allow for delegation. For instance, users should be able to say that they want to subscribe to mute and block lists from other individuals, groups or organizations they trust. One example of this might be that I could choose to automatically block anyone who is blocked by more than x% of the people I follow (where I can choose x). Ideally these features could be implemented at the tweet/conversation level and not just the account level.

So you can see that by decentralizing the power to the edges of the network INSTEAD of further concentrating it by requiring the network owner to further centralize power is the right answer, both from a technology perspective and a regulatory/policy perspective.

Sadly, I think we are in a race with ourselves in this centralization vs decentralization debate. We need the decentralized tech stack to evolve more quickly and show the world how decentralized technology works in a mainstream way at scale before policy makers and regulators force the tech sector to go the wrong way.

And, most disturbingly, the regulators and elected officials are taking actions, well intended of course, to slow the decentralized sector down, not speed it up.

Which is why we at USV have been spending a lot of time with public servants of all kinds, educating them, imploring them, and desperately trying to get them to understand where we are, why it is an important moment, and why we need to this new technology to succeed.


Comments (Archived):

  1. William Mougayar

    Indeed, one of the biggest threats to decentralization is current regulation that fails to see the new paradigm shifts it brings, and rather prefer to see everything with the existing lens they have.The 2nd challenge to decentralization is that – from a technology perspective, the “decentralization stack” is not well defined, so I will echo your remark there.And there are many ways to start chipping at (or getting to) decentralization. It is a layering / de-layering process. You need to be somewhat distributed before you become decentralized, and sometimes these 2 concepts are confused with one another.The stone age didn’t end for a lack of stones, so we can’t expect the central monopolies to die off before forging ahead with decentralization. Decentralization must thrive on its own, and leave the old central constructs that don’t make sense anymore behind.

    1. JamesHRH

      Stones were decentralized.;-)

  2. Pointsandfigures

    Great points. I agree. Regulation should only be used to fix negative externalities that cannot be fixed by the market or competition. In the short term, decentralized systems haven’t had enough time to develop to be meaningful competition. They are mere blips on the radar screen like some UFO.At the same time, Twitter, Facebook, and Google are bringing it on themselves. Instead of being even handed, the perception is they are being totally biased in the way they manage their user communities. When the head management of Google have a company wide mourning party after a Presidential election given the power they exert over the information that gets distributed and how it gets distributed, we have a problem.I like Albert’s thoughts and would hope that the companies police themselves before Senator Mark Warner and the rest of Congress sink their teeth in. Ironically, the head of the FCC might be a defender of the companies right to police themselves instead of cooperating with Congress.

    1. Pointsandfigures

      Just found this post that is counter to Fred’s thoughts on Blockchain.””We fared no better when we reached out directly to several blockchain firms, via email, phone, and in person. Not one was willing to share data on program results, MERL [monitoring, evaluation, research and learning] processes, or adaptive management for potential scale-up. Despite all the hype about how blockchain will bring unheralded transparency to processes and operations in low-trust environments, the industry is itself opaque.”Burg was an enthusiastic advocate for blockchain until recently – as he explained in this Medium post.”Blockchain is like a loom that can weave together multiple strands of separate things… into an integrated fabric where you can see what the data means and adjust resources in response,” he swooned.Blockchain has been wildly mis-sold, but underneath it is a database with performance and scalability issues and a lot of baggage. Any claim made for blockchain could be made for databases, or simply publishing contractual or transactional data gathered in another form” https://www.theregister.co….

      1. sigmaalgebra

        > “Any claim made for blockchain could be made for databases, or simply publishing contractual or transactional data gathered in another form”That’s what I’ve thought.Besides, IIRC, distributed has been a solid, standard feature, option of relational database software for a long time. Once I looked at it and got the impression that really mostly just have a central log file.

      2. Adam Sher

        You left out the money shot! 0.00% success

    2. PhilipSugar

      I am surprised they allow politics in the workplace. I don’t. But you are right the power is a problem. Me controlling speech in my house is very different than doing it on a massive scale.

    3. awaldstein

      Actually not a problem at all. Ill advised I agree. Everyone has biases, including you my friend, and the culture of a company follows the biases and beliefs of the leader.If you don’t believe in climate change don’t work for Patagonia.

      1. Pointsandfigures

        No, Patagonia has competitors. Google, FB, Twitter do not. Or, if they do they are marginal at best. There was no mourning meeting at Google the day after Obama was elected President-yet it was a very sad day for Conservatives.

        1. awaldstein

          Simply disagree.People believe what they believe and we establish rules to create some fairness and order around them.That friction is the stuff of life.Your logic that Twitter’s or Facebooks leaders don’t create rules that establish an attempt at unbiase even though they themselves are, doesn’t hold for me.We will never agree on this.(And I never think about this till you kvetch about it 😉

        2. LE

          Fwiw I truly despise many things that Patagonia says and does however I still buy their clothing and I wear it. LIke every day. A good brand. If it was easy to duplicate the particular items that they have with others I might switch but I am not going to spend any time on it either.

        3. LE

          There was no mourning meeting at Google the day after Obama was electedMany reasons for that (such as your point) however will note that conservatives (perhaps let’s hypothesize anecdotal) are typically more ‘grown up’ and rational and not emotional (in that particular way) to begin with.You know sometimes a shitty event happens that is on the news. And the emotional types (which can include conservatives for sure in some cases) are all over the place saying silly shit based on one event. As if the world is coming to an end. Or the action shows a pattern that indicates some larger problem (which is never the case). They over react. They are emotional. A rational person seeing the same event steps back and considers the scope of what happens and is more likely to be thinking clearly and rationally about the event rather than knee jerk off in some direction with an oversupply of feeling that quite frankly clouds their thinking. It’s very childlike actually. Social media separately exacerbates this phenomena.

      2. JLM

        .Say I am a brilliant accountant. Top of the profession. Up to speed on my FASBs and a CPA. I live for debits. I adore credits. I drool over goodwill impairment and non-cash compensation.A. I go to work with a heart which beats in tune with Al Gore’s as it relates to global warming, climate change.B. I go to work with a heart which beats in tune with the NRA and am skeptical about climate change.C. I got to work with a heart that lives for Alabama football and nothing else.I am A — a supporter; or, B — a skeptic, or C — an agnostic.How does this impact my work? Am I less knowledgeable about goodwill impairment, do my credits lose balance with my debits.When I am hired, who asks me these questions?Isn’t this the essence of a hostile work climate?Is part of diversity diversity of thought?If the belief in global warming/climate change is a condition of employment, then there is a storm coming and it will end up in a Federal Court.Global warming/climate change has become a religion and this is just religious persecution, no?JLMwww.themusingsofthebigredca…

        1. LE

          How does this impact my work? Am I less knowledgeable about goodwill impairment, do my credits lose balance with my debits.Breaking it down to the simplest way to look at it, the views ‘B’ (unfortunately) are (apparently) rejected by the typical person working at the company driven by the leadership (which shaped the culture down the line) . As such you (‘B’) won’t be able to interact and play well with others because they be biased or quite possibly even will hate you for the way you look at the world. That isn’t right and shouldn’t happen. But it’s a reality. The anger will impact your interactions with co-workers. That’s the reality.So from a practical sense it seems to make sense what is going on here.Now you could argue that the company can and should take steps to correct the bias and make the workforce accepting of different points of view.But think of some things that have been brainwashed into all of us and how hard it is to reverse those thoughts. That is what it’s like. The bell can’t be unrung with a meeting, a directive, or even training.

        2. Girish Mehta

          “I drool over goodwill impairment”Ha !!

    4. JLM

      .The Google post-election pity party was alarming on a number of different planes — such incredible sophomoric weakness, such obvious bias, and all in the face of a commitment to diversity.For Google diversity means you can look different, but, baby, you cannot dare to think different because we have a single way of looking at things and if it doesn’t turn out that way, then we hold a company wide pity party and you are not welcome.It is silly.JLMwww.themusingsofthebigredca…

      1. someone

        You should try academia.

  3. kidmercury

    Surely it is apparent from the slide above that decentralization simply unlocks value that is captured by a new centralized entity. The same will be true of crypto. This does not mean it is of no value, but it does mean that those who constantly cite decentralization as the solution are largely incorrect in the sense that *re-centralization* is the real outcome, as the slide above clearly illustrates.

    1. Holly Hester-Reilly

      That’s what I suspect as well.

    2. JamesHRH

      You missed the words ‘a fraction of the’ before ‘value’ in the first sentence.

  4. Matt Zagaja

    I think user configuration of software is “old school”. People stopped sitting down and adjusting their GMail spam filters long ago. The new paradigm is machine learning/AI driven configuration. Software will change its interface and behavior by learning from your interactions with it. Facebook and others already do this. The battle is whether this happens locally or on the server. The folks at Apple are betting that this is going to happen in a decentralized manner on your own device. Google and Amazon are betting that they can build better algorithms by aggregating everyone’s interactions in the cloud.

    1. JLM

      .Brilliant comment.JLMwww.themusingsofthebigredca…

    2. JamesHRH

      They are both right. Different value props need different answers.

    3. sigmaalgebra

      What I’ve seen so far about AI/ML guessing what I will like, am interested in, etc. flops: The guesses are inaccurate, insulting, intrusive, and patronizing.Maybe I’m just old school, but when I swing a hammer at a nail I don’t want the hammer to balk, refuse to act, suggest a different tool, make me get “elevated privileges”, and force me to confirm that I want to hit the nail. That stuff is insubordinate back talk!If that is the best AI/ML can do, then it will be a passing fad for enhanced user experience.

  5. iggyfanlo

    Could it be that the equity markets are early (as usual) iare telling us that decentralization is coming soon…In discounting the long term value of the FAANGs and presaging the almost inevitable dissolution of these tech monopolies

  6. Tom Labus

    You’re also asking that entrenched pols give up the cash they get to decide winners or tilt and stifle a market.

    1. JLM

      .Politics is a cesspool of money made even worse by the short period of time it is effective–the short pot life.It mines the conflict between two, hopefully, well-funded alternatives — the energy industry v the climate/environmental hustle for example.But what it does is live off this money in conflict.Whoever has the bigger pot of money will ultimately win. Maybe not forever, but as long as they have a deeper pocketbook.In many instances, there is no necessity to have a winner-take-all outcome.You could do some good stuff on emissions while building the Keystone Pipeline. The world wouldn’t end, but the politicians would have two sources of money to mine.Right now the established order has an edge–they are reliable political funders.The disruptors, by definition, both have lesser resources and less broad contacts. Those K Street boys have spent decades building their rolodexi.If you are looking for the politicians to discipline the selection of winners and losers based on right v wrong, good luck.If politicians see their role as making order from chaos, then they will always support more chaos.I think that is where we are, unfortunately.JLMwww.themusingsofthebigredca…

      1. sigmaalgebra

        Starts to answer a lot of questions I’ve had. For (1) always look for the hidden agenda and (2) follow the money; you filled in a lot of details.I have begun to suspect, as you explained, that (i) there have been some big, established funders pulling the strings as in the cover art for The Godfather, and they have been totally PISSED off at Trump (a) funding his own campaign, (b) flying his own gorgeous 757, and (c) getting votes quite directly from the middle and laboring classes in the “fly over” states.So, somehow these string pullers, feeling threatened, got the anti-Trump, 24 x 7, gang up pile on, get with the “narrative”, character assassination, repeat it often enough and people will believe it (or at least keep their heads down against the clique and tribe), ABC, CBS, CNN, MSNBC, NBC, NYT, WaPo.But the anti-Trump effort is having problems:(1) Trump may be partly right that it is an advantage to have Nasty Nancy back in power because then the Dems can carry some responsibility for some of what happens and could be hurt if things go bad.(2) Trump is darned good with the media, e.g., giving fast news conferences before each helicopter/Air Force 1 departure/arrival, Twitter direct to voters, continued rallies, LOTS of photo-ops.(3) The two biggies are peace and prosperity, and Trump has done well on both.(4) The 2+ years of anti-Trump attacks totally without substance and in strong contradiction to his accomplishments are starting to wear thin.(5) The Greenies crying wolf for so many years with their predictions now past and wildly wrong, no big, obvious problems from the Dakota pipelines, what the Greenies and Governor Moonbeam did to the California forests, the lower gas prices, lower unemployment numbers, faster GDP growth rates, continued low interest rates that people like are hurting the Democrats.IMHO it appears that some of Trump’s biggest problems are Republicans who liked Paul Ryan. Net, without Trump and all his effective campaigning, we’d have Hillary now. That is, without Trump, the Republicans are losers — Obama did let them win lots of elections but didn’t give Hillary a sufficient boost to win against Trump although I’d guess she would have won against any of the other Republicans.

  7. sigmaalgebra

    Good to read your reasoning about decentralization.Sure, decentralization is an option at various levels of the whole stack, all the way up to anti-monopoly issues, but on your points I’m “reticent. Yes, I’m reticent.”First, at the technical level, distributed is usually just much harder.Second, again at the technical level, for many years, people have gone distributed when it was really important. E.g., once I got a tour of the NASDAQ computer site in Trumbull, CT, and they explained how they have a second site that could take over, with little or no delay, loss, etc. whenever the CT site suddenly, totally crashed.Third, I’m not so impressed with the monopoly history or argument.For IBM, I was there when they were making their big decisions on PCs. IMHO, in short, IBM wanted PCs like they wanted, well something bad and ugly. Feet locked in concrete, they didn’t want PCs. Their favorites were CICS, IMS, DB2, JES, and their software wonder JCL. They didn’t really want their System 38 either.IBM wanted to be their branch offices with blue suits selling to big customers. Mostly can’t have those blue suits selling PCs.For more on the IBM monopoly, could plug together an IBM mainframe with an Amdahl or, IIRC, Fujitsu CPU, EMC disk subsystem, 3270 terminal clones (with an adapter card, a PC), but, right, it was still an IBM mainframe!Yup, big, bad Big Blue: I arrived on campus as a starting assistant prof, at the first faculty meeting heard the sad report from the computing committee, and stood and made a good proposal in a few words. A year later, my proposal was running. IBM tried hard — meeting with faculty for lunch in the Faculty Club, meeting with me a few times, offered us a mainframe for free, sent in their squad from their local branch office, had the head of the central mainframe shop strongly on their side and a sworn enemy of me, and sent in their national supersalesman Buck Rodgers. I beat them. Some monopoly.For the Microsoft monopoly, for my startup I considered both Linux and Windows. I picked Windows and continue to believe I made the best choice for me. Maybe if my business becomes a really big thing, then I’ll hire a systems programming staff of 100+ Linux experts and “do a switch over and hope I don’t get a burn out”! Maybe. I prefer to rely on powerful Microsoft (apparently a few days ago the most valuable company, past Apple by some for a while), partly because they are powerful. I doubt they will try to compete with me.I know; I know; Intel is a big monopoly with the microprocessor business all locked up. For memory, their Micron has all that. Except for my first server I bought an AMD FX-8350 — really impressive so far. Hurry up and buy one or several before the price falls farther, now $100-, quantity 1 retail. For the 16 GB of ECC DDR3 CS 9 main memory, it was from, I’m guessing, South Korea. Some monopoly.For Twitter? How about the top 2% of users with the most followers get together and start their own version?For more, someone could write some Twitter-like software, with improvements, and sell it, say, like Disqus. Then small subject matter verticals could use that software.Ah, right!! The verticals could share, be distributed, and look like one thingy!! Then they would merge and no longer be distributed!!!Google search? IMHO, Bing is about as good. There’s a lot I don’t like about Google along with a lot I do like. If Google shut down, I’d use Bing. Occasionally I use Bing now.Gmail? When my last computer was sick, for a while I used Gmail. I hated it. I’m back to Microsoft Office 2003 and its Outlook and like it. For a while I used the e-mail software I wrote long ago, but now commonly there is so much use of attachments with so many tricky encoding standards, e.g., quoted-printable, that getting a common e-mail message back to just simple 8-bit ASCII or UTF-8 text the way God intended is a pain; Outlook, even the 2003 version, does a good job on that. IIRC there is an API to the Outlook PST files, and I’d like to use it, but have other projects higher on my TODO list.Net, I’m not afraid of the big, bad monopoly monsters.

    1. Richard

      Anyone can be the owner of a publicly traded monopoly, at anytime and now at no cost other than the cost ownership.

  8. Girish Mehta

    Centralization vs Decentralization is a false trade-off.There is no “versus” there. You need some centralization for governance, for efficiency, for recourse. And you need to challenge “lock-in” for the next wave of innovation to emerge and flourish.Every hero needs a villain. Since decentralization is a “feature” of crypto, it appears that centralization shall have to step up to play the villain ?Then, there is the small matter of the customer – A large-scale, end-user-driven, legal demand for crypto.

    1. JLM

      .One lawyer in a small town? Poor guy starves to death.Two lawyers in a small town? They both live in the biggest homes in town.Competition creates riches.JLMwww.themusingsofthebigredca…

      1. Girish Mehta

        That’s funny.

      2. Sid

        Funny. I will use it 🙂

    2. LE

      Every hero needs a villain. Since decentralization is a “feature” of crypto, it appears that centralization shall have to step up to play the villain ?Who makes it the villain? Not the end user or customer but the investors. Isn’t that obvious?Users and customers don’t think or care about any of this. Investors and those (startups) that want a piece of the pie care about this. So they make it an issue under the guise of ‘better for the market’.Customers just want a need fulfilled. And they are happy even paying more for it as well. If the flight to the West Coast is $200 but could be $180 they are still happy at $200 because it seems fair for what they are getting. How has electricity dereguation gone? I don’t even get contacted by people wanting to compete for my utility cost (was the rage years ago) because most people don’t care about the competition and nominal savings. Not even worth their time to switch.Nobody cares that Starbucks has a lock on selling their ‘sugar delivery system’ by virtue of their size and market power. And ability to lock up prime real estate at less of a cost than some other person who wants to compete with them.

      1. Girish Mehta

        A lot of the “big picture” around crypto is made up by investors and startups to suit their narrative. End-users and reality are conspicuous by their absence.Sapiens was one of the best books of 2015, but I think we will look back to this time and wonder if the tech community got a bit carried away with the story of stories and narratives.

        1. sigmaalgebra

          Recently heard that the idea of “narratives” — e.g., for PR, propaganda, building fads, blowing bubbles, having an endless chain of media stories easy to write, political attacks, etc. — was pushed, maybe dreamed up, by the ad guy Bernays.

      2. sigmaalgebra

        > Those who do not know history’s mistakes are doomed to repeat them.The pile of moonbeam BS you cut through, shoveled into the Atlantic, would raise sea levels by some feet!

  9. JLM

    .These kind of discussions are very useful because they let free ideas which NEED to wrestle to become bigger, stronger ideas. They also expose huge voids which have to be filled.The big issue with platform v publisher is, believe it or not, legal.A platform can argue that it is not an editor of content and therefore not responsible for what is contained on its platform. The “willful blindness” defense.A publisher has certain first amendment protections — being able to print stolen material as an example, talking to you Wikileaks and Julian Assange, the NYT, the WashPo — but must discipline and police the content of its site as it is liable for the truthfulness of statements.The law of defamation (libel, slander) is not complicated, but you have to break a sweat to prove you’re disciplining your own team.The tradeoff becomes one of protections v liability.Small example, but Fred Wilson has some responsibility for what is posted on his blog. One may say it starts at zero and ends with complete responsibility, but it is something. Is he a publisher or a platform? Not taking any stand as to what it is, but it is something. Unfair? Yes, I think so.On the issue of censorship, Twitter and others are in a real pickle. There is enormous evidence — starting with the direct testimony of Jack Dorsey — that the organization puts its thumb on the nature of content both explicitly and accidentally.When you are dead, the fact that you were killed by an accident is not much comfort.Dorsey said that “conservatives” working at Twitter were fearful of speaking up. That is a Hell of an admission, first by the founder/CEO, but also from the perspective of creating a hostile work environment.Going out on a limb here, but I don’t think corp legal cleared that utterance.The exact quote was: “…I mean, we have a lot of conservative-leaning folks in the company as well, and to be honest, they don’t feel safe to express their opinions at the company.”https://www.dailydot.com/de…Throw in the recent political brouhaha in which the Russians stole the rightful election of Hillary Clinton from her talons [please tell me you get this is sarcasm, OK?] and you have an arena which is toxic when it comes to politics even before any of the participants piss into the punchbowl.I am aghast at Twitter’s admission that they took ads from Russians which were paid for in Ruble denominated credit cards. [Pro tip: If you are concerned with the Russians meddling in your elections, start looking at the ads paid for in RUBLES.]Bottom line: There will never be a pure decentralized solution to anything, but there will be varying degrees of centralization (regulation locus) and de-centralization. It is like drinking absinthe. Just enough to hallucinate, but not enough to blind you.JLMwww.themusingsofthebigredca…

    1. someone

      You are interesting.

      1. JamesHRH

        Not news.

      2. Some girl

        No , he is sexy

  10. Todd Lash

    i am a believer in decentralization and taking power away from monopolies. mesh and p2p networks for internet access, power generation, and many others will be a huge benefit for society. I don’t know how to reconcile this with a decentralized “facebook like service” that is rife with fake news and hate speech but no throat to choke.

  11. Richard

    Modern Day crypto VCs – Let’s confuse politicians make up imports sounding terms like “Decentralization” / cryprocurrentices, $$$ “Talk To Policy Makers”, it’s for the good of the people.

  12. jason wright

    Is it fair to say that regulators are susceptible to the will of the most powerful vested interest over the wider general interest of society? Regulators seem to come from law firms, firms with corporate clients, clients with a clear vested interest in influencing the direction of new regulation to serve their ‘needs’.

  13. LE

    That is how elected officials and regulators often think. They look backwards to find a model of regulation that has worked in the past and try to apply it to a new thing.Fwiw ‘they look backwards’ linguistically infers a negative and is a put down in this context. May be better to say ‘elected officials often believe that what has worked well in the past is therefore appropriate for the current situation or the future’.

  14. Ron Shah

    Crypto has created a capital market. That is it’s crowning achievement.Decentralization is still just a theory. The pioneers of internet never sat around hoping the public sector would help them. They never stood on a soapbox about the benefits. They made it all happen, and people used it for all different types of things, and it grew. No one could stop it.This is the major defining difference. This is happening with crypto as a way to store and transfer capital but not much beyond that. There is nothing, in sight, happening on a grassroots level that makes its mainstream aspirations real (that I know).You know much more than me about this so would love to understand if I’m missing something

  15. Holly Hester-Reilly

    I’m intrigued but not convinced that it is so simple. Yes, regulators are overwhelmingly behind the times and confused when it comes to tech. Their current day actions do more to entrench than foster competition.On the other hand, I don’t see how mass decentralization solves the problems we have today. In theory perhaps one imagines each actor in the system making good choices for themselves, but haven’t our latest economics taught us that that is not how societies function?What if instead, moving to decentralized smart contracts only makes those with the tech talent and knowledge even more in control than they already are by reducing the government’s ability to take action as decided by the people?I don’t claim to be expert enough to know what will happen, but I have real doubts about crypto as the solution.

  16. Frank W. Miller

    If I had a nickel for every time I heard someone claim that we just need to decentralize and it will fix everything…Your chart is fairly inaccurate as well. IBM’s mainframes did at one point have a near monopoly but that was not replaced by open source. DEC came along with the PDP-11/VAX computers and took over because they were smaller. Intel did the same thing with x86 because they were smaller. This had nothing to do with open source.WRT software, I think you should check the stats on OS usage. Windows is still completely dominant. Linux may be used in the server space but the desktop is totally not open source. iOS is definitely not open-source and I would argue that Android itself isn’t either.And I would differ with you on FAANG going away as a result of “open-source” crypto networks.You’re missing a fundamental tenet here that really has nothing to do with the tech itself. We’ve known how to make fully distributed networks for a long time, long before the crypto variant came along. The problem is exactly what you hold up as its strength, the lack of control. Somebody has to pay for these things. And if someone has to pay the only way they’re going to do that is if they can make money. The only way they can make money is if they can control things.Those data centers cost a lot of money, real estate, electricity, etc. While you might make the argument that my phone can act as a small chunk of that and the summation of all the world’s phones replace those datacenters, the real problem is, no one can make any money on it. The Internet existed for a couple of decades until somebody figured out how to make money on it using the web which is when it really took off. You can’t make money (in the traditional business sense of charging for a product or service) with crypto unless somebody tries to control elements of the fully distributed system design. But as I’ve said before, that flies in the face of the very essence of their design.I think full distribution is only good for allowing academic computer scientists to get their tenure… 😉

    1. Holly Hester-Reilly

      Well put!

    2. Pete Griffiths

      Agreed.The slide is a major simplification of what happened. A fairer representation would be1 Mainframe dominated by IBM with plug compatible but competitors financially viable due to the system lock-in2 major strategic to that was dominance mini computers with DEC leading the way3 the critical steps that changed the world to be more open were UNIX and PCs although they were different in kind4 UNIX was adopted by practically all mini computer companies with the specific intent of creating a critical mass of mid range application Software. UNIX was open source. Various licensing SNAFUs led to UNIX snatching defeat from the jaws of victory but not before it spawned LINUX which went in to dominate p what became the server market.5 PC/Wintel dominated the desktop. Finally crushing even the UNIX based technical workstations eg Sun Apollo DEC HP. Not open source but easily implemented by multiple vendors.

      1. Lawrence Brass

        Still not open source but “publicly licensed”, where do you put ARM and its domination of the mobile landscape? Can it be 6?

        1. Pete Griffiths

          Good point about Arm.

    3. Asim Aslam

      Disagree. I think the whole reason this works is because blockchain has unlocked token economics and the ability to build trustless systems. We will build decentralised networks with token economic models which layer upon the existing centralised systems to provide new value.https://github.com/micro/ne

      1. Sebastian Wain

        The problem is that we need to see yet a single use case that works. Currently there are serious limitations in scaling dapps. If you look closely these limitations are “solved” partially centralizing the system.

        1. Asim Aslam

          Dapps built on ethereum execute on the Ethereum virtual machine which is known to be slow. They lag behind. Look at advancements in this space. Running the WASM virtual machine is by far the best and most performant way to run these applications. Blockchain itself is just for the token economics. There should not be any on chain execution of code.Finality in terms of transactions can also be sped up. Seed Dfinity’s approach with random beacon, BLS and threshold relay. I come from the world of distributed systems and cloud computing. The problems everyone is trying to solve are already solved, its just a case of understanding the right way to partition and horizontally scale. Dfinity has taken the best approach so far.

          1. Sebastian Wain

            > The problems everyone is trying to solve are already solved,Not clear where it is solved. BFT-SMaRT The top BFT library has a performance of 80,000 tps for just 4 nodes and degrades substantially when you add more nodes. This performance is not acceptable for dapps.

    4. Paulo Trezentos

      Regarding the slide, everyone of us can read past facts in a different way. I think the slide summarizes well the forces that drove change. Was DEC important? Of course, and Minix, Stallman, Posix, Bill Joy, …

    5. Matt A. Myers

      Your comment above ties well into why I believe, if blockchain is to be used as a true currency – perhaps means to bring all economies together, and there’s true and immutable accountability, so we put ourselves literally all in the same via economics – it will be by each country and their tax base paying for these costs including distributing mining in all participating nations; and there’s no chance, other than by manipulation or force, that a society would or should support the costs if it’s structured like a Ponzi-Pyramid scheme and where it isn’t elected officials in place as part of control mechanisms.

      1. janasotkov

        mattamyers important

    6. lonnylot

      I don’t think it’s a stretch to say that iOS and Android are based off of open source Linux/BSD and that mobile has become the dominant player, not desktop. Whether iOS or Android are open source are irrelevant.

    7. Julien Boyreau

      @@fwmiller:disqus , probably the greatest summary of why Business Bounded Full Decentralization is an oxymoron. Thanks.

    8. Michael Elling

      I would add to your and Pete’s view that it is also not full stack and doesn’t take into account a lot of other exogenous factors.Much of the change in layers 3-7 and edge/core compute push/pull dynamic is a function of changes in layers 1-2 and also what was happening in other verticals (voice, video, etc…). By Layers 1-2 I mean equal access in 1984 which paved the way for the commercial foundations of the internet, followed by cable-broadband and wireless 7×24 access. The latter only happened after multiple, rapid, evolutions of technology and ultimately scaled due to a regulatory loophole from the 1930s; aka part 15 and WiFi.(Note: Steve Jobs single-handedly resurrected equal access in 2007, which the Powell FCC had killed in the early to mid 2000s. Otherwise we wouldn’t be where we are. Another “exogenous” event that no one talks about but is fundamental to what occurred in the chart.)But all networks follow the same rules about layer 1 to 3 tradeoffs (vertical) and edge vs core processing (horizontal). Analog and digital. Real and virtual. Technical and socio-economic and political. I really wish the crypto crowd would go back and bone up on the other aspects of their education beyond STEM.And inspected more closely, it is a very narrow and misleading perspective. After all, didn’t company #2 on the chart close on Friday with the largest market cap on the planet followed closely by #3? And #1 transitioned from a network effects compute model to a linear consulting model. It never figured out another geometric value model. Hence it’s relative, not absolute, decline. And IBM appears not to have learned even when it comes to blockchain: http://bit.ly/2U9O287Garbage in, garbage out.

    9. alexmyers83

      For now, maybe – but centralization of this kind cannot scale forever since data cannot survive as capital forever. There are myriad reasons for this, so the paradigm has to shift. Data as labor would take it’s place, and the value of individual user contributions will create the resources needed for compute/storage/etc. https://blog.singularitynet… discusses this new paradigm in some detail, and must be the way forward.

  17. george

    Conceptually, I don’t think these are new ideas for decentralizing a platform (Ref reading: Platform Revolution) but nevertheless, I agree with the concept or principle – that consumers can create more value (multi-user feedback loop) and potentially achieve this by design -modularity.POV: At the core – the Twitter topic really isn’t about decentralization or centralization – I believe it’s more about governance and compliance. However, I do believe this is a byproduct of decentralized thinking – most of the large platforms that have scaled over the past 2X decades haven’t address these issues because phylosophically, it’s really not part of their decentralized core vales (exception Apple). Inevitably, if its not part of your core values, chances are, it wasn’t originally built into your core interactions on the platform.The past few years have been amazing, new business opportunities have emerged in this area – developing subsystems (modular approach), relying on both owner and user to help improve the platform experience – it’s more self-reinforcing that way.

  18. Pablo Yabo

    Twitter is a good use case that can work decentralized if we focus just in public accounts. Most of the other platforms have to advance to cover privacy issues of decentralized platforms.

  19. Grace Schroeder

    In the case of security tokens vs. traditional securities – the existing regulations are part of what gives the US the largest % of public companies. Interesting, too, that the technology —- if allowed — does as good or better a job as some of the intermediaries vis a vis adherence to the compliance rules associated with jurisdiction and class of security. The rub is that there is no regulated party that the SEC can hold accountable if, for example, a poorly written smart contrsct/whitelist allows the exchange of securities in violation of the regulations. I believe that strong collaboration with technology providers (not just lawyers) would accelerate resolutore. Tokenization of securities makes too much sense to take a “wait and see” approach and to forfeit our leadership.

  20. Asim Aslam

    Very interesting post and quite timely. I was having a discussion with a friend yesterday about the cyclical nature of all systems in nature including the ones we build ourselves. It is very clear that after we built a decentralised network called the internet and then build centralised services on top of it which consolidate data we would then build a decentralised layer on top. All layers evolve at their own pace, all layers must thrive for the ones above to grow and become the foundation for the next cyclical event. Decentralised networks built in the coming decades will also become the foundation for the next centralised systems but they will likely look very different, dare I say, be autonomous agents which we have given control to.Dfinity is taking one approach towards decentralised compute but I feel as though there needs to be a higher level focus. Cloud providers like AWS have given us powerful centralised compute, edge networks like CloudFlare are extended that reach and so I believe there’s an opportunity to build a micro service layer on top.The work begins now https://github.com/micro/ne

  21. Emil Sotirov

    Crypto is like early Christianity – a necessary anarchistic impulse against imperial structures. The context – Romans invented the concept of History (as we understand it today) because they needed a way to think about and manage highly asynchronous processes in a highly distributed environment. “History” became the universalizing abstraction of the Empire – above the multitude and variety of local provincial ideologies (the myths and religions we call “pre-historic”).Christianity open-sourced the Roman “History” tech. Once Christian, everyone could start and manage their own copy of the “history” framework. Individuals could now have personal history, provinces could become kingdoms by having their own history. But here’s the thing – Christianity posited a public “master” copy of the History framework – God – as a truly abstract (symbolic) principle unifying all local installs – “under one God”. Important organizing aspect of the Christian “God” abstraction was the fear of punishment for bad management of a local “history” install. Now, “history” – as a modeling and management construct – “explains” things as a timeline of causes and effects. The “history” owner gets a timeline (aka “lifetime”) and a chance to “explain” him/herself and “become” (hopefully better) instead of being once and for all classified as “good” or “bad”. Our modern notion of personal “freedom” comes from this uniquely Christian mandate. Other variations of the single “God” abstraction were more strict – they would monitor and punish you right after you did something bad. Those variations were closer to another universalist “tech” invented by the Romans – the Roman Law.Anyway, if Crypto wants to “win” as Christianity did… it would need an even more universalist “God” abstraction capable of instilling appropriately corrective levels of “fear” in the billions of now ever freer “nodes”. Will we see new (post-imperial) “feudal” structures organizing into Crypto kingdoms? Can Bitcoin somehow do the job of a “God” abstraction above all – know-it-all/known-by-all, ultimate consensus on who did what, reward/punish all?

    1. JLM

      .Interesting, thoughtful insights.The Romans conquered foreign lands with their military expertise and might and not one foreign land that they conquered was ever assimilated. Every conquered land hated the Romans and their strictures.The Romans tried, sometimes successfully, to draft local soldiers and to teach them the Roman centurion formations.The Romans empowered tax collectors by specifying how much they had to send to Rome, not how much each local had to pay or what the tax rate was to be.The Roman tax collectors – locals – got to keep whatever they could get beyond the mandated tax number to be sent to Rome.The Christians were an underground religion in conflict with Rome, Judaism, and idol worship.The Christians, like other religions, tried to create a body of written word to capture their culture and their origin fable.JLMwww.themusingsofthebigredca…

    2. sigmaalgebra

      Wow.My ugrad school was really big on liberal education, e.g., trying to understand history, man, and religion — they had a big freshman, sophomore course “Man in the Light of History and Religion”. They tried. I could never find the beef in that supposedly elegant sandwich and, instead, rushed off to the departments of math and physics where I had a shot of telling truth from fiction.I still can’t hope to get a good grade on the history course discussion (sling BS) question “Why did the Roman Empire collapse?”.My best guess at the drivers and causes of history are technology and economic productivity, but I can’t claim that there isn’t more that has been powerful. For the technology, the standard claims are that the keys were agriculture, some productive grains, irrigation (the Egyptians), domesticated animals, writing (for remembering and communicating over long distances), arithmetic (for bookkeeping), geometry (for land surveying), and astronomy (for calendar construction and open ocean navigation).I do remember the warning “History doesn’t reveal its alternatives.”. So, in particular can’t do well controlled experiments and, thus, are essentially blocked from doing anything scientific. Or the Santayana “Those who do not know history’s mistakes are doomed to repeat them.” aside, the study of history, even worse than that of psychology or sociology, has little or no predictive value.

      1. Emil Sotirov

        I hope the “Wow” is not as in “Wow, what BS” … 🙂

        1. sigmaalgebra

          No, you are safe! I meant “WOW, you are REALLY good at writing about history.” What you wrote brought together and connected a lot of stuff and was at least interesting and plausible. I can’t do that with history; what I’d write would not be plausible or even interesting. So my ugrad school tried to get me to write plausible answers to questions about history, but I never learned and, instead, just rushed back to math and physics.Instead of history, it would have done me a lot more good to have had a course in human female behavior, motivations, and emotions! Couldn’t offer such a course now since the UN Politically Correct Police would arrest the prof!

  22. Pascal Aschwanden

    It’s awesome that you’re trying to educate politicians on the right moves to make. I hope the US doesn’t end up like the EU, as far as their efforts to regulate the web. Ironically, the EU does a better job of regulating their broadband providers, at least for now.

  23. JamesHRH

    This is not a tech ramp. It is a chronology.The need was computing. The solutions just moved parts of the value chain up & down the stack ( data, middleware, UI ) and actoss the hardware chain ( servers : routers : devices ).Users don’t care where things are, they just care that things get easier or better.Crypto-philes think a product attribute is a market need.Classic mistake writ humongous.

    1. sigmaalgebra

      Hate to say something so direct, but IMHO excellent summary and conclusion.Maybe Fred/USV see a theme, paradigm, major quite broad and general force of synergy via all of (i) an end to monopolies, (ii) open standards, source, and APIs, (iii) distributed systems with security, ownership, etc. implemented with (iv) crypto, etc. I can’t prove this is just moonbeam stuff, but I have a tough time believing in it; to me it is a long way from what finance takes seriously.I look for things more specific and solid, evaluated one at a time, and don’t see much utility in prognostication of paradigms.Maybe Fred is waiting for some confirming, specific, actual good business opportunity pitch decks to arrive in his e-mail. Maybe.I believe it really is possible to see 5-10 years ahead in some important, useful ways but only rarely, under special circumstances, with some special, powerful tools at hand, and only quite specifically and likely narrowly.

  24. Bob Walters

    > From the slide: “Cryptonetworks –> open source data”What is a cryptonetwork and how does it lead to open source data? Wait…lemme guess…a blockchain! I wish that people would be more precise with language. Here’s a news flash: crypto already runs the world. And blockchain runs almost nothing. This wholesale substitution of “crypto” for “blockchain” is disingenuous at best and self-serving at worst. It reminds me of substituting “climate change” for “global warming” (or for “pollute less”).> “…regulators…are taking actions to slow the decentralization sector down…”I thought that regulators were taking actions to slow the centralized sector down (through antitrust, privacy provisions, etc.). Wouldn’t this provide fertile ground for the (undefined) decentralization sector. Oh wait, you mean the blockchain sector? And ICOs?This whole blog post makes me uneasy. The inaccuracies and omissions are one thing, but the fact that VCs now have the power, wealth, and inclination to become lobbyists for failing investment theses by “…spending a lot of time with public servants of all kinds, educating them, imploring them, and desperately trying to get them to understand where we are, why it is an important moment, and why we need to this new technology to succeed…” makes me ill.These attempts to tilt the regulatory playing field are the stuff that undoes free markets.

  25. Lee Blaylock

    This is the kind of perspective and deep thinking I really appreciate Fred sharing. This is a great blog/community and I appreciate it. Thank you.

  26. Terry J Leach

    Two years ago I pitched my cybersecurity startup to several VCs as example of how to get policy makers to understand the benefits of decentralization, but they did not understand. Timing is everything and I don’t give up easily, today I’m still working on business and technology. In 2019 we will have the prototype in beta testing, we can talk about adding our story to your pitch deck.

  27. riemannzeta

    The fundamental problem we are going to have making this case to regulators is that the same decentralization trends that will break up for-profit monopolies have potential to break up not-for-profit monopolies (a/k/a sovereignties).I believe they get that at the gut level even if they don’t understand the details.

  28. Lawrence Brass

    Wow.. take my money, Gregory! :)Open source today is not what it used to be. You see a lot of contributions to open source coming from and financed by private corporations. Retro-feed happening at all levels, technical and project governance. IBM’s Redhat acquisition, Microsoft building its services heavily on Linux, big data corps built upon OSS. It is not linear. More likely an outward exploding spiral.

  29. Asim Aslam

    I think something like this is a network that allows you to collaboratively build software from existing services in the network. It cannot be built by one person alone, it must be a collaborative effort. If you had 1000 engineers working on building a service on top of decentralised network vs facebook’s own centralised engineering efforts then you’d have the ability to compete. It requires a network that behaves like the internal platforms that these companies have. People are getting lost in the world of decentralised compute. That’s not where the value is, the value is in the ability to build services.https://github.com/micro/ne

  30. Gregory Magarshak

    Well if you are serious about it, contact me: greg at the domain qbix.comI would like to see preselling of utility tokens being used to finance open source projects eventually. But there is much more that can be done. Take a look at qbix.com

  31. Gregory Magarshak

    Exactly. There needs to be an ecosystem similar to what we have in WordPress plugins, for instance, but perhaps even more robust.