The Emerging Architecture Of Internet Applications

The bitcoin blockchain is not just going to change the way money works on the Internet (and off). It’s going to change the way Internet applications are built. We have been working hard to understand how things are going to look in five to ten years and Joel Monegro has been providing much of that thought leadership inside our firm.

Since we are not into keeping our insights to ourselves, we have encouraged Joel to publish all of our work in this area (and every area). And today Joel has posted something that is really important and needs to be understood by every Internet/mobile entrepreneur, investor, developer, employee, and analyst. It is the blockchain stack and it looks like this.

blockchain stack

The most important things to understand about this blockchain stack are the overlay networks (most of which are still emerging), and the shared data layer and the shared protocol layer. Please read Joel’s post which describes each of these in some detail.

What is most important about this emerging stack is, in Joel’s words,

This imposes a very interesting set of challenges for developers, entrepreneurs, and investors as so much of the value in the current Internet stack will be commoditized by this architecture.

Differentiation and defensibility and network effects will be much harder to obtain with this architecture. Most things will work like email. Take your keys from one app to another and all your data and relationships come with it.

Fun times are ahead. Time to put your seat belt on.

#entrepreneurship#marketplaces#VC & Technology

Comments (Archived):

  1. pointsnfigures

    The best defense (and sustainability) will be scale and figuring out how to monetize around that scale.

  2. William Mougayar

    Where on that stack are you focusing your investments?

  3. William Mougayar

    But since the premise of all this is Decentralization & distributed technologies, could we depict this using a lateral / horizontal paradigm, and not a vertical one?

    1. Guest

      Is this what you meant?I wonder why the Overlay Network has two boxes and same for the Decentralized Protocol, 3 as supposed to just one cutting across? … and Perhaps, App, App, App, App, should just be Apps.

      1. William Mougayar

        I think overlaying examples would help. Joel hinted he might do that.

  4. Twain Twain

    Can Joel make this graphic less developer-y (less vertical stacks) and easier to understand for the layman?Try a fusion of a cross-section cut of the Earth where the inner core is where the BTC miners are, the crust is the overlay network (where Coinbase seems to be positioning itself?) and then there are spokes going off (like in a classic Neural Net Open Graph) to the APIs that then node over to different Applications.

    1. William Mougayar

      Yup. That’s what I meant in my comment about being horizontal, not vertical. There’s another user friendly version that can be depicted.

    2. Joel Monegro

      Thanks for your feedback. I agree that there are better ways to represent this.

      1. Twain Twain

        Something like this…..Hyper-distributed data with BTC protocol in the bake, spanning all nodes in the Web.My diagram’s an annotation on top of National Science Foundation-University of Michigan research into Facebook data memes.*

      2. Simon Edhouse

        I think the diagram is entirely appropriate, for the reason stated below, but also, the ‘technology stack’ in the software industry has a lot of history and logic behind its vertical layout.

      3. Amar

        @jmonegro:disqus A possible option would be to show an example implementation. I.e. have the exact stack/block diagram but replace abstract problem-domain names with actual implementation instance name. So <<xyz>> on top of <<xyz>> on top of “lazooz” on top of “sidechain” on top of “blockchain” on top of <<…>> (i picked these from reading your well written post :-), thank you!)I agree with @twaintwain:disqus .. any work you can do in moving the conversation simultaneously in both the abstract and concrete domains will be very helpful. Otherwise you lose a lot of your initial audience rapidly.

      4. Twain Twain

        Citi Chief Economist says, “Gold is a commodity like bitcoin and other crypto currencies”:*

  5. falicon

    I appreciate the thinking here but believe the assumptions (and hence predictions based on them) are completely wrong.There simply isn’t enough motivation or benefit to the various players to make this a reality…

    1. Steve_Dodd

      my simple question is “where’s the money”? If the foundation is a shared data layer, how is it to be sustained (paid for) to deliver the value upwards?

      1. pointsnfigures

        advertising….heh Seriously they will have to figure out a way to monetize eyeballs, data and other functions around the core product. Medical companies could give the drugs away for almost free, and monetize the ecosystem around the data.

      2. Simon Edhouse

        money is now shared data… the value can enter this schema at various points, but is underwritten or given integrity by bitcoin.

        1. Steve_Dodd

          I can understand that but who’s paying for the data (or access to it) and how? Or, who’s setting the price (Bitcoin or otherwise)? Companies that provide the infrastructure to deliver the platform so users can create “data” seem to be ignored here. The value of Advertising seems to be at risk as advertisers are challenging the value being received and consumers are becoming alienated by the ads they are seeing………..

          1. Simon Edhouse

            I repeat.. money, now = data. So, “who’s paying for the data” is a non starter as a question. The ‘Platform’ is not being provided by Companies is the other aspect.. and ‘Advertising’? not sure of your point… But ‘Advertising’ is another old stale relic that absolutely needs to be transformed, and become much more user centered. (hopefully that will flow from some of these fundamental changes)

          2. falicon

            What companies are directly profiting from ‘data’ right now? Data is powerful and useful, but it’s not nearly as *valuable* as everyone would like you to believe.Right now – the value is in how you use data more than simply having data (everyone has data; everyone can get data).Advertisers are waking up to this reality as well (which is why everyone is panicked about “where is the money going to come from?”)

          3. Simon Edhouse

            “..directly profiting from” is a subset of “how you use data”, I think.. and which companies? The two mega examples that come to mind are Google and Facebook… which both, (by the way) make obscene amounts of money from data freely given by their users.But data in this context is not what I was referring to previously.. That kind of data will likely come from apps on top of the ‘blockchain application stack’.. the idea of money being data is that for the first time, we have a reliable mechanism for monetary exchange that is so close to open source internet stuff as to be part of that meme… and absolutely not part of previous money memes.That’s why folks are getting so damn excited.

          4. Steve_Dodd

            This really helps! One thing people always wrestle with is the definition and categorization of data and types. Perhaps there is a better way to define the “shared data layer” within this specific use case.

          5. falicon

            Google and Facebook don’t make money from ‘selling’ their data though…that was my point, people think they will just collect/own a bunch of data and then that builds some magical value. Google started by collecting public and available data…but it wasn’t useful until they applied their search aglos. to it (and started generating data on top of that…and using that data to drive more products/features…etc. etc. etc.). They aren’t valuable because of the data they collect, they are valuable because of the users they attract and the way that they let users benefit from the data….splitting hairs here really, but I think saying data is valuable is still a very dangerous assumption out in the general mindset.The money aspect of BitCoin and the Blockchain is specifically what I think makes it all meaningful…and yet ultimately limited in it’s application/adoption.It absolutely solves a few of the problems around money store and money exchange…but so much of the current internet is only loosely tied to these sorts of transactions (and almost *all* occur as an aside, behind the scenes, part of the interaction — at no ‘real obvious’ cost to the end user).If I have to pay a micro fee every time I post a comment, share a photo, or click a link (or worse every time one of these things is pushed *to* me)…it’s going to end up adding up and that’s going to fundamentally change the habit of consumers (not nec. a bad thing – but likely an unrealistic shift compared to what we have now).I’m not saying the blockchain ain’t cool…it is…but I don’t think it’s gotta be shoved into and under everything…it makes lot more sense to me as a very specific tool that you plug into certain situations/apps…but not everything.

          6. Steve_Dodd

            Got it, thanks!!!!

          7. ShanaC

            how do you mean?

          8. Dave W Baldwin

            Advertising will become focused on prediction due to data processed in real time matched with a more consumer friendly system.

        2. pointsnfigures

          interesting point about money=data, have to think about that. we currently monetize data…..

          1. Simon Edhouse

            obviously monetizing that kind of data continues.. ~ Saying that money (bitcoin) is data, is just a way to get folks to realize that Bitcoin is different.

      3. Jeff Jarvis

        There’s another question: Where’s the money saved? Where’s the open-source dividend? Where is efficiency created? Where is friction eliminated? Where are middlemen murdered? Don’t just follow money creation. Follow money savings.

        1. Salt Shaker

          Aren’t trust and security large motivators? I think those who shop at Target, Neiman Marcus, UPS, Goodwill, P.F. Chang’s, Michaels and Home Depot have seen friction in the current system.

        2. Richard

          Growth and Saving are alike in many ways, each marginal increase becomes successively harder.

    2. Simon Edhouse

      ‘Disruption’ is a badly overused term.. to the point where it has become trivialized and lost its true meaning. Lots of ventures want to be ‘disruptive’, its the new ‘cool’… But its not that simple. Real (and massive) disruptive innovations take time. Also history has shown us that they are usually dismissed by the vast majority of people who see no relevance in something that doesn’t apply to their immediate daily lives.But that should tell us something.. i.e. If any disruption is really worth getting excited about, it will almost certainly seem quite unrelated to all the ordinary things we know. That’s what Bitcoin and Blockchain technologies represent. They are related instead to the things that most people do not know.Or.. let’s put it another way. ~ You can’t judge future ‘supply’, based on current ‘demand’.

      1. falicon

        I agree – but in this case part of the problem is that Joel starts his post freely admitting “We spend a lot of time looking at everything that’s going on with Bitcoin and the Blockchain”.This is the clasic case of preaching to the choir…what Fred probably *should have* said is that this is what the future of the Blockchain world likely looks like…not *the* future.

        1. Simon Edhouse

          There are reasons why such bold predictions are being made.. Because when macro changes take place (every 10-20 years) they tend to sweep everything away like a tsunami.. and lord knows some tired old systems deserve to be shown the door.

          1. falicon

            And yet people still write letters, read books, listen to the radio, and watch tv…the more the world changes, the more it stays the same :-)People will always want to connect, tell stories, and share experiences…we’ll continue to find new/interesting ways to do it…and *most* still won’t care about the ‘how’ of it all (only those of us trying to figure out ways to profit from it all really will).

          2. Simon Edhouse

            People still practice witchcraft, (or think they do) and in Nigeria cannibalism is still rife.. Its a big wide weird world. ~ But I think we are talking about macro trends.

          3. Chimpwithcans

            Rife is a strong word – existent maybe – but rife?? Spent my whole life in Africa and never seen one person eat another.

          4. Chimpwithcans

            OK – that makes it look pretty rife. Is this the image you guys in USA get given of Africa?? I can testify that the whole continent is a lot nicer than that video. Not a bleeding heart in sight πŸ™‚

          5. Jeff Jarvis

            Yes, people still read books. If you’d like to build a future business based on that, more power to you.

          6. falicon

            If they get authors like you involved they’ll be just fine… ;-)p.s. Really enjoyed the “Gutenberg the Geek”

          7. Richard

            Of course, but maybe there is something else going on? 60% of books that are purchased online or off are not read (at least to completion). 40% of college graduates never read a book again after graduation. Show me a great book and you could publish it on toilet paper and people would read it given if it were marketed correctly. What if instead of the weekly presidential address (which about a dozen people listen to) , the president presided over a book of the week club.

    3. Uriel Fanelli

      The point is, over a certain limiti, you don’t need the players to do anything. Bitcoin works perfectly with local dealers for small business, so in any case, either the players will be convinced, or the players will be disrupted.

    4. andyidsinga

      hey kevin, seems to me if the stack ends up like the diagram above, it will be useable by apps similar to the way ssl and https is today, and probably better security. what do you think?this also make me wonder if investment returns here will be lack luster even though the technology is fundamentalcheers!

      1. falicon

        Yes – if this comes to be I think you are right on most of that. I’m not sure if it would be *more* secure than SSL is, but it could def. be more ‘trusted’ than SSL is today (in fact I think that’s the core thing the BlockChain addresses, trust without the need for the 3rd party).I don’t know of any efforts to replace SSL certs with BlockChain technology yet…but *that* is an interesting idea…hrm… πŸ™‚

  6. JimHirshfield

    Great to see more plain language posts about the protocol stack. Tipping Tuesday worthy.+300 bits

  7. Evan Van Ness

    Fred, serious question: how can you write posts like these and not even mention Ethereum?

  8. Jorge M. Torres

    Centralization and control are powerful incentives for people to invest money and time into new projects. Joel says that 80% of the emerging stack is occupied by open and decentralized technologies. I wonder if the remaining 20% represents enough room to generate returns sufficient to justify the investments needed to radically transform the way people experience Internet applications. Is this something people even question/debate nowadays?

    1. ZekeV

      If you own bitcoin, then you benefit from the open source development efforts, without having to build a pay gate around any particular business.

      1. Jorge M. Torres

        Sure, but what about incentives to build the few parts of the stack that can be gated? I suppose if the market opportunity is really, really, big, then it will be more than enough.

        1. ZekeV

          Proprietary software attached to a commercial API can be monetized in the traditional manner, I suppose. If it’s useful, people will pay for it. For example, let’s say Blockstream implements its sidechain specification for a proprietary platform that processes micropayments. Nifty! Surely they can make some money on this. But, they would generate a lot LOT more wealth simply by purchasing and holding a large stash of bitcoin.

        2. ShanaC

          how big is the market?



  9. LIAD

    Wow! You guys actually do research and analysis and think things through?I thought VC’s just threw money around based on how cool an entrepreneurs haircut was.

    1. LE

      You guys actually do research and analysis and think things through?I thought VC’s just threw money aroundIt’s important to be able to justify what you are doing by making people think that you are earning the money that you are making. So what you do is create the appearance of “research theater” for people who invest in you and people who write about you. You claim that it’s part of the “secret sauce” that allows you to be successful. You aren’t going to say “we are just throwing darts at a board”. [1]The above is said partly in jest. I don’t understand enough about what the research paper says to have an opinion on the value of it. However the part about “be able to justify” is actually a very important concept and is separate from whether any of this research really has any value at all.[1] A variation of this was on display last night when the DA talked about the Michael Brown case. I watched the entire speech. He didn’t just get up and say “Wilson won’t be indicted”. He took a very long time to present the facts presented to the grand jury in order to convince people that “hard work had been done and this is fair”. And hard work had been done. But as with my other point it was important to do this regardless of whatever the truth was in order to stave off any criticism or lessen it. So the length of the discussion of the case served a purpose of “here is why this is fair”.

  10. LIAD

    if you control the shared protocol layer you control the rest.the devil in me says the best way to exploit blockchain technology is to co-opt the core developers.Anyone got a spare horse head?

  11. Greg Kieser

    I personally think blockchain layer will be multiple blockchains – not just bitcoin – and we’ll see decentralized apps straight to the top of the stack.Seems likely the greatest innovations of the blockchain technology (eg. app and api layer) will be achieved by disparate groups of developers incented by appreciating appcoins as opposed to venture backed entities.



  12. Steven Roussey

    This reminds me of BlockCypher.

  13. Elia Freedman

    there has been a lot of angst over Twitter’s moves lately. It would be interesting to analyze those in relation to this stack and see where they overlay/go off the rails.

  14. taylorwc

    From Joel’s post: “Protocols are hard for the average developer to build on top of, so there’s a[n] opportunity in making it easy to connect them.”This! How many services have we seen crop up that abstract complexity away from the user? It’s an awesome trend that can be seen in so many different layers of the existing web stack.

  15. Simon de la Rouviere

    Also worth taking a note at Ethereum who wants to build a similar setup:

  16. andrewparker

    I read though Joel’s post. Interesting read.The architecture he describes is useful for web services that suffer from the Byzintine Generals problem. If you don’t trust Twitter to not mess with the timestamp of your tweet or eBay to not mess with the clearing price of your bid, then building on top of the blockchain will allow you to trust no one while simultaneously feel confident that all “transactions” mark in the network are accurate an validated.I’m not sure that’s really the next phase of the internet. Or, if it is the next phase of the internet, then that means our existing traditional internet has suffered the tragedy of the commons and all traditional internet services are reduced to a lowest common denominator of distrust.If you do continue to trust the web service provider you use for a given transaction (Twitter or eBay in my above examples), then I think this new proposal for a completely decentralized internet stack suffers from failing Occum’s Razor. It’s overly complicated *if* you generally trust your counter party.

    1. David Semeria

      Comparing Diaspora and Facebook’s user stats highlights the difference between words and actions. If people were so genuinely concerned about privacy, the subscriber numbers would be inverted.So I don’t think there’s much upside in promoting the blockchain to the masses on a trust ticket. However, I think the opposite is true when it comes to companies. As Fred once said:Don’t be a Google Bitch, don’t be a Facebook Bitch, and Don’t be a Twitter Bitch. Be your own Bitch.Open API’s on top of the blockchain would be everybody’s bitch – and therefore no-one’s.I believe identity and authentication are the most obvious places to start.

    2. Joel Monegro

      Actually, I don’t think this is about trust – it’s about decreasing transaction costs and reduced take rates.Trust certainly wouldn’t be the driving force of this stack. I generally don’t have a problem trusting Twitter or eBay, and I’m most people don’t care about whether a service is centralized or decentralized. I do, however, care about cost and I think most people think the same way.If we look at the ride sharing market as an example, the vast majority of users just go for whoever is cheapest when they need a ride. But companies like Uber and Lyft have cost structures that prevent them from going below a certain point.An Uber competitor built on top of this stack (or a suite of them sharing the same protocol and data) could operate without a take rate (making them consistently 5-10% both cheaper for riders and more profitable for drivers) and that’s something Uber – or any of the many companies we have today that are very profitable at X% take rate but wouldn’t be at X-1% – can’t compete with.

      1. andrewparker

        What is it about this stack that leads you to the conclusion a web service built on top of it can operate without a take rate? Uber’s infrastructure and data hosting costs are trivial compared to their human capital needs (thousands of FTEs, not including 1099-issued driver fleet) to run their business. Without a take rate, how would they change to manage this operation? Perhaps you’re heading for the conclusion that it can all be decentralized, like the DNS of ride sharing πŸ™‚ ? I am very skeptical that would ever be logistically possible. A business is so much more than the information architecture on which it’s built.

        1. Joel Monegro

          Admittedly that was a bit of an exaggeration to make a point – the protocols themselves enable zero-take rate transactions but the companies on top of them will have to take X%.But the general idea is that decentralization allows for X to be vastly lower than the 10-30% (and sometimes even 50%) many companies take today, and that these companies would be unsustainable at X% and therefore unable to compete.What’s especially interesting is, going back to the ride sharing example, that if you decided to skip the company and connect to the protocol directly you could transact without a take rate.And regarding the DNS of ride sharing, you should look at what is doing πŸ™‚ If you built an Uber competitor on top of that protocol you could add a couple of proprietary value-add services on top (reputation, safety guarantees, etc.) but it would be a much thinner layer than the entire Uber stack, so you could get away with a much lower take rate.

  17. Brandon Burns

    As the technology becomes commoditized, design, experience and branding will be the defining features of products. As has always happened.Auto engineers in the early 1900s were specialized laborers. They’re still specialized, but it’s definitely a lower paid, assembly-line job now.Printing in the 1800s and before was done by specialists who both physically printed materials and did typography, layout and other design elements since they were inextricably linked to the press itself. Now the actual printing is 99% done by machines, while design is still done by specialists are paid make printed materials standout.But for “everyday” web development as a professional practice, as more folks learn to code, will look more and more like a trade school skill, as it now is with auto engineering, or even start to become more and more obsolete, like in printing (however, there will always be room for technology specialists and those doing the groundbreaking work, and they will always be highly sought after and compensated). And as the need for web dev sills decreases because due to commoditization, people will look for a differentiator β€” and they will remember that the need for good design and branding that speaks to people and touches them emotionally will always be something that technology cannot and will not replace.

    1. LE

      Printing in the 1800s and before was done by specialists who both physically printed materials and did typography, layout and other design elements since they were inextricably linked to the press itself. Now the actual printing is 99% done by machinesPrinting presses still require specialized operators. The ones that take printing ink that is. Many parts are computerized and automated but printing presses involve feeding paper through a machine and laying ink on paper and there are many many things that a human still needs to do to keep one of those running correctly.Below a video for paper feeding on a smaller press. This is just paper feeding there is also ink water balance and a host of mechanical issues to keep in check.…This is stuff that I’ve actually done btw. Also typesetting and graphic design (on light tables and right when it got computerized in the 80’s). It looks easy in the same way playing a musical instrument looks easy if you’ve never tried to actually play something. (Or tennis or skiing etc.)

    2. LE

      But for “everyday” web development as a professional practice, as more folks learn to code, will look more and more like a trade school skill, as it now is with auto engineering, or even start to become more and more obsoleteI’m not seeing the comparison between web development and auto engineering. Web development takes creativity even if you have the technical skills.Being a bit older than you I remember when the mac first came out and people were able to do their own brochures and design their own collateral material. Those with no design training ended up doing things that were “ransom note grade” in quality. A total mismatch of fonts, line weight and so on. Many things looked really really bad. Some were hideous. That said the people creating didn’t even know enough about design to think they looked bad (probably because it was their creation like not seeing your own baby is ugly..)

      1. Brandon Burns

        I think the analogy to auto engineering is one about commoditization due to streamlined processes and access to standardized tools. If you previously had to make the components for an engine and experiment with all the different ways to build one, but then things change to where the process of making an engine is standardized, decreasing time spent doing it by orders of magnitude, and on top of it you can buy all the components in bulk, then the amount of labor required to make the engine is decreased, and the kind of knowledge required becomes less rare, which also makes it more easily accessible and cheaper to hire.The same goes for building web products. A good example is Pinterest. It was a unique idea once upon a time, but now social photo bookmarking platforms are more numerous. So numerous in fact that, if you take One Month Rails course, you learn in 30 days how to build a fully functioning Pinterest. It probably took several folks months to build the original Pinterest but, now, all of the sudden, the ability to build Pinterest is put in the hands of way more people, and the skill becomes devalued. And part of the reason why you can learn to build it so quickly is because the course gives you all these 3rd party tools, from plugins to existing code libraries, that do things automatically that one used to have to hard code from scratch β€” which is the root of Fred’s argument of “taking your keys from one app to another” and how the mere existence of “keys” makes it so that you not only spend less time trying to figure out how to “open the door” but more people can do it, commoditizing the skill and the labor force itself.Granted, there will always be room for folks doing things that are actually new. But for the “everyday” things, Fred’s right that the technology becomes commoditized. I’m making the argument that the people and skills required to build that technology, in effect, become commoditized, too.

        1. LE

          Ok all correct. You said “auto engineering” but my brain thinks of this as “auto manufacturing”. Engineering to me are the guys with horn rim glasses in the 60’s at mission control in houston. White shirts.Auto manufacturing is a wonder of the dropping costs of mass production and computerization. You get tremendous value for what you pay when you think about it.One big difference though between the auto business and the web business is “iteration”. The big benefit of the things you mentioned above is the fact that it allows for really quick iteration and trying many different things most importantly by many different people.In the auto business there aren’t many different people in the game trying and able to wack some idea together over a weekend and see if it works. And if not, try something else. And all info at your fingertips.In my first business I actually had to rent space to even see if the idea would work. I had to find a guy to come in to help me learn how a machine worked. It took time to find that person and for him to be able to fit it into his schedule. Today that would be a web video away same day learning. In another business it took me a year to get together all the info that I needed to know what equipment to buy I had to get on airplanes to see the equipment. (I took my own videos which I still have this was about 1994/5). Most of what I spent my time on I could easily do today by never leaving my desk. So in that one year I might been able to try 5 different things vs. 1 thing.

          1. Brandon Burns

            I think that’s a great point, about engineering vs. manufacturing. When new technology emerges, its all about engineering. But as the engineering aspects become less needed due to commoditization, there’s a transition from engineering a thing to manufacturing it.I’m gonna steal that from you. πŸ™‚

          2. LE

            Glad that it’s good enough to steal.I love factories machines and manufacturing. I love the sounds that equipment makes. I can still remember the sound of the stepper motors in the brand new phototypesetter that I bought (Itek) and the sound of the sucker cups on the paper feeder on the presses. At the end of a good day you’d want to wipe that machine down the same way that someone wants to brush a dog. Down to the gleaming chrome. Very satisfying. Lot’s of intermittent reinforcement ups and downs.

    3. falicon

      devs. have def. been living in a ‘feast’ environment for the past 20 years or so…it won’t (and shouldn’t) last forever…just like with music, the world and the value will *eventually* return to focusing on true artists.I’m not sure that any of this is helped or hurt by the blockchain though (if anything it probably extends the life/need of devs, as this sh*t is still pretty complex and confusing — meanwhile traditional web and mobile worlds have finally started to become reasonable for ‘regular’ people to kinda build)

  18. William Mougayar

    Now Mario Batali can fully understand the Blockchain architecture. You can tell him it’s just like layers of a birthday cake.

  19. Kassey

    “Take your keys from one app to another and all your data and relationships come with it.” This is amazing! Now we’ll have the freedom to move in and out of a walled garden.

  20. PrometheeFeu

    Dont take this the wrong way, but I can’t decide whether your focus on blockchains is foresight or madness. Blockchains depend upon their guarantees holding forever. If the guarantees are found to be flawed in 20 years, everything unravels. And you can’t just switch algorithm because now all your history is in question. This seems to me to be a very dangerous foundation.

    1. falicon

      Fred can afford to be a Blockchain/BitCoin champion – there’s HUGE upside (for us all) if it works out and very little downside for him personally if it doesn’t.The problem is that we’re going to lose a lot of lemmings in this march…

      1. PrometheeFeu

        My point is that it’s a bet that could hurt if it succeeds for a long time before failing. Imagine a world where blockchains are pervasive. Your money is bitcoins, the deed to your house is a colored bitcoin, so is the stock in your portfolio, etc… Now some guy comes along and discovers a flaw in the algorithm which allows faking transactions. You can switch algorithm for future transactions but the old ones are no longer trustworthy. The whole chain is instantly compromised. So you dont have anything to transact with anymore. The algorithm is a single point of failure with cascading catastrophic consequences if it is found wanting. I think that is too-fragile a system and so as much as I love the whole crypto-currency concept, here, I hope it fails.

        1. FAKE GRIMLOCK


  21. zvozin

    Bitcoin mining has a deeply fundamental security issue built in by design, which makes it vulnerable to collusion. At present it can at best be partially mitigated, but still requires a degree of social cohesion in the community to guard against that may or may not be possible to enact and maintain.EDIT: to short-circuit possible technical arguments, the attack in question is _not_ the Sybil attack, and is a larger, meaner and more plausible version of what’s been discussed at….Explanatory blog: http://hackingdistributed.c….Dry-as-old-bones paper:

  22. ShanaC

    This looks like current web architecture, stack upon stack. Web architecture also could be designed in other ways – as a network structure, which bitcoin can’t be as mining gets harderThe thing is, bitcoins, while decentralized, are open to security attacks, as mentioned below.This doesn’t make them good money. because money needs really networked structures to stay liquid and spendable to be a good money.

  23. James Arsenault

    Forgive a half-finished thought here, but the comment that network effects will be much harder to obtain didn’t quite make sense to me. I suppose traditional closed system network effects will be harder to obtain (e.g. critical user mass on a traditional social network like twitter).However, let’s imagine what twitter would look like if it were designed in five years. By then, ideas like OneName will have taken off. You own your identity (real or anonymous), and it is propagated through the blockchain. You’ve associated your identity with your browser (or phone more likely), so you don’t do much logging in to anything – unless you say otherwise, your online activity is associated with the identity you’ve selected.So, when it becomes apparent that the world needs a place to remark on things 140 char at a time, you’re not really thinking about how you’ll build a large enough social network to achieve minimum threshold network effects. OneName (or something like it) has evolved and all the social connections and valences that used to be held on facebook and google+ live in the blockchain, governed by a social protocol. And whether users are on twitter or facebook or elsewhere, they can easily leverage their social network. Imagine being able to @mention someone across the breadth of the entire internet regardless of which medium you were in!!I think the network effects will be built into the internet, leaving MOST companies/organizations to focus on striving for differentiation and defensibly (and quality!) instead. Rather than focusing on building a sustainable user base, twitter would be focusing on how to make sense of how to leverage the topography (eg. relationships and histories of interactions between users on other web properties) of the general social network to find meaning and relevance in the conversations that take place.The exception of course being among competing protocols, which will require certain scale and popularity. But the whole sidechain concept can help with that a bit as well.TL;DR: Blockchain makes network effects easier, not harder. The tough part is attracting the attention/focus/dollars of the already-established network…

  24. falicon

    Also want to nitpick/rant on this bit for a second: “Most things will work like email. Take your keys from one app to another and all your data and relationships come with it.”I think I get what you are saying here…I can create a message in one system/app and other systems/apps will be able to receive and digest it…that’s the whole ‘like email’ bit right?Have you tried to move email providers recently? It’s not any easier than it was years ago (especially if you are trying to keep your email address active and alive). Your data and relationships don’t come with you.If you change your email client, and you’re using a centralized email server, then sure your stuff comes with it…but that’s not any different than if you were using Twitter web, and you start using Twitter mobile…*wow*, my tweets and everyone I follow is still there!Besides the point is that the blockchain doesn’t change the ability for any of this from the app point of view…if I want you to be able to export/import your data from one system into my system I can do that now (the trouble is very few of the app providers *want* you to be able to do this)…Sure if we agree on the blockchain as *the* standard place (and we agree on a standard format for ‘data’ within the blockchain)…then it’s easier to write the imports and new companies/apps can bootstrap quicker with that approach (but again, existing apps kinda don’t want that — though you could argue facebook, twitter, and even google *already* give you this).But these are some REALLY big ifs…and at the end of it, if it all comes to play out, where are we? Same sorts of apps and features…with millions of dev. hours (and dead startup attempts) laying in the wake of our path to get here…That doesn’t make sense to me.If this revolution is going to happen it’s got to be sparked by something more than the promise of “same but better behind the scenes”…it has to offer up something we can’t (reasonably) get any other way AND that we all *must* have…please show me that. I want to be sold.

  25. Sriram Yadavalli

    Stack makes sense, but I dont understand why a company like Uber would do transactions via blockchain or even put any data there. Data is their biggest asset and using blockchain would commoditize them



      1. Chimpwithcans

        A model for pure accountability? Uber definitely not interested in that πŸ™‚



  27. Gregory Magarshak

    True… I see this as disrupting the VC industry to some extent, because there will be less advantage to being a first mover, since there won’t be any vendor lock-in. Do you have any thoughts as to how the investor community will adapt to this?

    1. fredwilson




  29. BerlinAgency


  30. Donna Brewington White

    Heard a pitch at TechWeek LA from a company (Tether) that is using the bitcoin blockchain to move dollars. It was eyeopening to realize that the blockchain extends beyond bitcoin.

  31. jason wright

    stack and blockchain fit?

  32. falicon

    On train (and heading into tunnel) so not yet – but in a bit

  33. JimHirshfield

    “isn’t appropriate” as in NSFW?

  34. Simon Edhouse

    ..some things are best described as being built on top of other things, (in such a vertical array) because they enable, or are enabled by them.

  35. William Mougayar

    yup, I can appreciate that software engineers think in terms of vertical stacks, but I think a non-technical functional depiction would be interesting.

  36. JimHirshfield

    ideal? representative?

  37. William Mougayar

    Oculus VR app πŸ˜‰ with smart watch notifications each time something new adds to it.

  38. Twain Twain

    Haha, indeedy.We do know Ray Kurzweil of Google’s already working on the brain chip nanobot that connects with the Cloud, right?*…It’ll help us think and remember, apparently.

  39. falicon

    OK – here’s the quick gut reaction thoughts:Thesis: The decentralized, blockchain driven approach is what *could* happen, and you could probably even argue that it’s what *should* happen…but I’m still pretty sure it’s not what *will* happen. It’s the BetaMax of this debate (in the sense that it’s the superior solution; but likely the market loser).Attempted Explanation: Step back and think about it from the higher level for a second…what specifically does this new approach bring to ‘the people’ that they don’t already get or have? Who are the big companies or services that are backing/promoting/pushing this approach? Who are the paradigm shifting upstarts that are using this approach to build something people never even knew they wanted or needed?When parts of those questions can start to be answered with specifics…then *maybe* this future will have a shot…but until then, on the higher level it’s more hope and dream than reality.Now let’s jump down into the lower levels and talk about motivations for specific parties.At the end of the day I believe this is all about connecting two parties through technology…usually those parties are not in the same physical location and often without the technology would not have any access to — or maybe even knowledge of — the other parties.In the current system making that connection actually requires a massive amount of middle-men (internet providers, servers, conversation facilitators, etc. etc. etc.)…but to the user, the conversation is pretty easy and mostly, often, feels one-to-one. It’s fairly easy, fairly cheap, and fairly fast…but there are still problems (fraud, asynchronous communication, cultural and political differences, etc. etc. etc.)…So there’s def. room for improvement…however I fail to see how most of what is being talked about here actually addresses any of these things in an important, big, useful way…sure Blockchain helps address the fraud problem, and that’s big enough to make it important and useful for the future…but not as *the* crucial part of the stack (in fact I think it’s more likely a piece of the stack that is rarely plugged in — especially since there is a micro transaction cost to the users when it *is* plugged in).All of this is basically my long-winded, rambling way of saying that the idea of a world without gatekeepers and middle-men is *awesome* and I hope it does come about…but there is not real incentive to any of the parties *doing* the work to make it so…and there’s no clear incentive for the users to demand it be so as well…Honestly, there aren’t even any real ‘gotta have’ services opening the eyes to users to this possibility yet…and I’m not sure anyone has an idea how or what will actually get there.

  40. gorbachev

    tl;drexisting architectures / services / solutions are good enough.That about sum it? πŸ™‚

  41. LE

    that the idea of a world without gatekeepers and middle-men is *awesome* and I hope it does come aboutGatekeepers and middlemen are awesome if you’re on the right side of what they do. They are a barrier to entry and if you are able to extract value from them then there is nothing not to like. The other side of the coin is the proverbial “race to the bottom” that can happen if there are no middlemen or less middlemen. Just like the other side of the coin with online reviews is merchants spend a large amount of time and energy catering to unreasonable squeaky wheel people.Fred’s a gatekeeper you know. And you can probably extract more value from Fred (over time) than you can from a system where there isn’t a Fred. And Fred isn’t even a gatekeeper that is directly related to you in any way. But you are definitely on the right side of Fred which is my point. If I was Charlie Chyrstal I’d want gatekeepers because I would know that after I social engineered the gatekeeper others would have a much harder time selling their “Secaucus 7 Seeded Bread” at LE’s local Whole Foods.That said like “the cloud” people have moving definitions of what a gatekeeper or a middlemen is to suit the particular point they are trying to make.

  42. falicon

    it’s not that they are good enough – it’s that the new approach needs to bring new value (not same old thing, different approach).

  43. Simon Edhouse

    maybe those dependencies are not so optional..

  44. falicon

    That’s the current sales pitch…I like it…but that’s a small part of my daily internet interaction…and even still I don’t know what it really means to me personally as a user (I don’t care how it gets done, just get it done — and hey, doesn’t facebook already sort of do this for me? Shouldn’t they? I’ve already invested a lot of time learning and building trust in that system….)

  45. Morgan Warstler

    The current margin / cut on Uber is 20-28%% – meaning Pareto, the low hanging fruit is gone. If it was Groupon at 50%, it’d be ripe.There’s still plenty of 80% Pareto low hanging fruit of things Uber can do – but chasing after them by trying to release the 20% ont he shit they do already isn’t the winner, IMO.Now replace Uber with “guys smart enough to get in in the middle” – getting in the middle is incredibly hard now, the rewards flow as such, that doesn’t mean that eventually nobody gets in middle.Basically, the Marx view of capitalism and perfect competition is wrong.

  46. Simon Edhouse

    some layers enable other layers, so the higher layers are dependent on layers that enable them.

  47. falicon

    But it sure feels like the general public does…

  48. LE

    If you want to get a good idea of what the general public cares about just invert anything that Hacker News cares about.To wit:”Postal Service almost never denies mail-surveillance requests”…Like I (the general public) didn’t even read that thread. Opened it up now just to link to it. Saw it appear for several days on HN. A non issue to me.

  49. Simon Edhouse

    I didn’t actually compare the blockchain with the internet.. but, I think its wrong to say that “the killer app of the internet was communication” because, firstly that isn’t an “app”, (so lets replace app with aspect) The killer aspect of the internet was ‘connectivity’ actually.. and communication is a thing that is enabled by connectivity. Its also important to realize that the blockchain is another thing that is built on the ‘connectivity’ of the internet, (TCP/IP) it doesn’t in any way replace it.The answers to your questions 1-5 could be summarized by ‘watch this space’.. The point of my first comment above was to say, hey, disruptions take time. ~ Look at ‘packet switching’ and how long that technology took to slowly be good enough for internet data transmission and then, only after many years of fast moving developments on the internet was it suitable for use with VOIP.The web commentariat has the patience of a small child.

  50. Thomas Petersen

    Here is a couple of ideas.Digital PokemΓ³n Cards. Basically the ability to create digital collectors items.E-books that could gain value. You could sell ebooks at a premium and allow people to re-sell them. Because the history of their ownership is recorded you could even see them gaining value if they had been owned by a celebrity.Private but public healthcare records for research and usage. Store the health-records publicly as personas but allow individuals people to link it with their identity. This would help with research in completely new ways.Voting made public but anonymous Make it impossible to fake voting results by making the results available for everyone.Artificial Intelligence Using the protocol to create automated consensus models.Companies without owners Build a company with a political purpose without any owner.

  51. BB

    For someone who actually believe that ‘cannibalism is rife on Nigeria’, your confidence in your knowledge of anything meaningful is quite impressive. Keep it up!

  52. ShanaC

    *sigh* sarcasm