The Web 3 Stack

Web 3 is the next generation of the web in which decentralized apps (dApps) operate on top of a shared data layer and users have control of their data and the ability to move between dApps with little to no switching costs.

Think about the way domains and email addresses work. We (individuals and/or companies) own these identifying data elements and we can provision them in any app we want (I provision my email addresses in gmail and my domains in wordpress but I could choose many other options). In Web 3, this is how all of our data will work.

But we have a long way to go to get there. The infrastructure for Web 3 is immature and at least a few years away from being mature and stable enough to build mission-critical dApps on. We can see glimpses of Web 3 in games and collectibles, where the stakes are not that high, and we can see glimpses of it in financial services that have built up around Bitcoin and other leading crypto assets. But it will be a while before we are keeping our resumes on DinkedIn and our documents on DDocs, our friends and family networks on Dacebook, our photos and videos on Dinstagram, and our President’s tweets on Dwitter. But that time will come. I am more sure about that than anything else right now. The last few years of Web 2 have shown us what is wrong with the current architecture of the web and what we need to do to fix it. When this all happens is another thing.

Our friend and colleague Kyle Samani posted a long and thorough description of the current state of the Web 3 stack yesterday. My colleagues Albert and Nick provided some feedback on it before he published it so it reflects some of our thinking too.

As investors, we are spending much of our time wrapping our heads around these infrastructure layers, trying to understand which approaches will win out, where value will accrue, and when dApps can be successfully built, scaled, and delivered to market.

We think this is where the action is right now and we are heads down in this mess trying to understand it and invest in it.


Comments (Archived):

  1. LIAD

    alongside the tech complexity which is enough to make your head spin, deciphering where value will accrue – at longterm equilibrium remains, at least to me, elusive.multiple theories, most diametrically opposed to each other. non-existent analogue/historical examples from which to draw comparisons/insight.continual tsunami of innovation which guts yesterday’s thinking and necessitates constant new analysis. – allure of VC to me previously was somewhat limited to the challenge of building a thesis, and then picking and assisting a handful of early stage cos from inception to hopeful the intellectual stimulation of figuring out this mega-revolution, with all its moving parts and complexities and then helping build the future from the ground-floor has super-charged that 100x.

    1. fredwilson

      It’s fucking hard

      1. LIAD

        BRING. IT. ON

      2. Twain Twain

        I’m looking towards the Chinese. There are protocols being designed and coded by Chinese engineers in AI, quantum computing and blockchain that are F***ING CLEVER.Every major paradigm shift needs the invention of a maths/science tool. In the case of Industrialization it was probability. In the case of Digitization it was Shannon et al. In the case of the next major leap, it’s know-how which the Chinese have had for 5000+ years.China’s sovereign fund is pouring money into quantum. https://uploads.disquscdn.c

        1. SubstrateUndertow

          Just maybe Shannon’s signal/noise channel analysis calls out the feasibility of stable quantum data/processing channels ?Every process has its hull speed limits.

          1. Twain Twain

            Ah, the Chinese are obviating Shannon.

      3. JamesHRH

        This is my worry for you – to adapt a quote from @andyswan ……. no one cares how hard it is, they only care what it does for them.Just cause you like solving hard things, doesn’t mean they are valuable……

      4. Twain Twain

        It’s simple. That Web3 stack design repeats all the same mathematical logic problems as Web1 and Web2.https://uploads.disquscdn.c

    2. LE

      alongside the tech complexity which is enough to make your head spinThe complexity is actually often a compliance technique. In other words you purposely write or present in a manner that makes it less likely that anyone will take the time to poke holes in what you are doing. Because it is written or presented in a way that is to hard to understand on a quick take. [1]There is also the tendency to put an intellectual halo around things that don’t make sense. So you infer that the person doing what you don’t understand must be smart and if they are smart they must be right. (Happens in art museums also with art as only one example).I don’t know enough about what is being said here to know if that is what is happening here.But on the other hand tech types have a clear history of making things way more complicated and obfuscated than they need to be and this, in most cases, seems to benefit them.A part of tech is built around ‘we are smart you are stupid’ so when something goes wrong with your shit PC an average user will think it is them and not the crap the company built that they are using. This only happens in tech. With, say, lawn mowers or a car it’s ‘fix your lousy product it doesn’t work right’.[1] Also the complexity justifies that it is brains and not luck that has lead to the success. ‘See what I do is hard this is what I need to know to make the money that I am requesting’.

  2. falicon

    Reading your list of future services made me go “Doh”…

  3. Vitomir Jevremovic

    Databases. Distributed, block-chained databases are one of the biggest missing links in creating Dapps today. Holding data in smart contracts or IPFS files is not a sustainable strategy. Of course we cannot have everything on the day one, but without decentralized databases there can be no true decentralization of Dapps. I might even say that block-chain should evolve into data-chain and that crypto currencies should move from ledger format into database format.. some day.

    1. Steven Roussey

      There are database-esque things out there in various states. However, decentralized means public, so best practice should include not putting highly sensitive data in a public ledger, even if encrypted (encryption algorithms get broken eventually, and while in a private server you can re-encrypt and jettison the backups, in the public chain everyone including bad actors made a backup for you, and are off decrypting your users’ SSNs some years from now — and no one will know since there will be no evidence of a hack). #decentralizedmeansquietglobalhacks

  4. William Mougayar

    I’m critical of that Web3 depiction. The diagram is mixing apples and oranges and makes no sense architecturally from a stack build-out view. Stack pieces should be complementary and wholesome to each other in an architecture. And doesn’t fit Ethereum’s nor Gavin Wood approaches, who were the original architects of blockchain web3. What is more likely needed is more standards and fewer upper layers, like what I described here: http://startupmanagement.or

    1. kidmercury

      i like kyle’s depiction, particularly as fodder for deeper thinking, though i think competing visions are likely to be worth exploring as the game is far from over. as i was thinking about it more this morning i thought that using government as an analogy may be helpful, in that the “stack” used by nation-states to create and execute their governance systems may be a useful corollary for thinking how blockchain stacks will evolve.

      1. William Mougayar

        This is not a stack. Stack pieces work with each other. It is a mumbo jumbo depiction of various ecosystem pieces, mixing protocols, middleware, functionality pieces, cloud capabilities, scaling approaches, content methods, etc… It’s not helpful in shedding clarity.

        1. kidmercury

          maybe it’s not a stack in the purest technical sense of the term. how about “architectural blueprint”

          1. William Mougayar

            it’s more like a technology list. blueprint may be too strong

        2. Gonzalo Arzuaga

          It’d be nice to see a visual image of a better stack, have you seen something like that posted elsewhere? Or it’s in your post you referenced above? Thx

          1. Jim Ritchie

            As a member of the Enterprise Ethereum Alliance I think our stack depiction makes more sense.…There are some very smart engineering minds working on this from major technology companies such as Microsoft and Intel.

          2. William Mougayar

            I do have anther diagram in my post.

        3. JamesHRH

          Whoa. Extra bite in today’s espresso?Psst – I agree.‘Shedding clarity’? Spreading maybe or some other lighting word.

    2. Lawrence Brass

      Blockchain web3 is an acceptable term for me.Plain “web3” is aspirational coming from the crypto crowd.I expect web3 to be greater and broader than crypto.

    3. Jim Ritchie

      Here is a video on an alternative stack depiction from the Enterprise Ethereum Alliance.…Graphic is in another comment in this thread.

  5. awaldstein

    Somehow I feel smarter when I listen to Kyle on podcasts then read his conceptualizations which of course, means i understand little.This is like a headspace mediation to me.

  6. jason wright

    For Nassim Taleb the only way to ‘predict’ (not is choice of word, but you get the idea) the future is to subtract what is here now that will not be in the future. What remains is the possible. There will be forks (and probably spoons).Mess 1.0 (which is opportunity time).A glossary would help.

  7. OurielOhayon

    the chart is missing a central part of the stack: the custody/wallet part

  8. Tom Labus

    Tough enough to grasp buy you guys have to put serious cash on the line too!!

  9. kidmercury

    i think blockchains are going to evolve to offer more and more of the components of the stack to dapp developers, so that developers essentially have a turnkey solution. however, i think expect each blockchain to prescribe the stack that it works with, so dapp developers do not and cannot make many of these choices themselves; or rather, the extent to which they have a choice is in which blockchain they build on. sort of like how android and iOS give you all sorts of APIs to facilitate mobile app development, but the entire stacks are different and thus to build an app on each one requires building an app compatible with the architecture of the given OS.

  10. Pointsandfigures

    I wonder how people are thinking about governance. It’s such a loaded word. Invisible hands can govern which should be possible with a transparency cryptocurrency etc. Governance might also be very light, with a committee elected to hear arbitration when misunderstandings occur.

  11. _miki_1774

    Seems very similar to Tim Berners Lee’s project Solid ( but built on top of blockchain. If the objective is to get back control of their data to users, is this complicated stack based on blockchain really necessary? Is there anything obvious that I am missing that makes this architecture superior to Solid? Everything is a nail for a hammer I guess.

  12. Kirsten Lambertsen

    If blockchain is going to power Web3, does the blockchain community look at the immense energy consumption of mining much the way we 1.0ers looked at bandwidth/delivery speeds? Like, ‘we know it sucks right now, but we also know it’s going to get to an acceptable level?’

    1. Lawrence Brass

      Cryptokitties CTO Dieter Shirley was very clear about the state of the art regarding performance and scalability when he was interviewed by Fred at the Token Summit.

      1. Kirsten Lambertsen

        I’ll have to go back and rewatch that one. Thx 🙂

          1. Kirsten Lambertsen

            Thank you!

  13. Frank W. Miller

    With respect, this explanation is fairly light on detail. It appears to be a mishmash of pseudo technical elements with very little actual technical detail and marketing hype. I was particularly taken aback at the “tens of trillions” and “world’s leading protocol designers” references. I see very little here that you could base any kind of design or even business activity on in this description.

  14. Sebastian Wain

    Great summary from Multicoin Capital. They can also take into account Snow White as another “leader-based blockchain consensus”. The stack diagram needs much more thought to be assembled.

  15. JLM

    .In the crawl, walk, run manner in which tech and the Internet have developed, it may be presumptuous to suggest that the transition from Web 2.0 to Web 3.0 is accomplished by wholesale abandonment of what has laid the foundation and the embrace of blockchain as a new foundation.It is, after all, a database and is already being absorbed by traditional users as part of their Web 2.0 existence. In this manner, it is just another “incremental” technology rather than one that is so fundamental as to replace the foundation of the web.Maybe blockchain is just another app which will be part of, but not a big driver of, Web 3.0.We are still awaiting the killer app which will remove the scales from our eyes and provide us with a clear vision of the future.Pretty clear that the promise of Bitcoin and crypto currency, in general, failed to rise to its potential and is now regulated like any other asset by the same regulators everyone hoped to avoid. Why is its cousin, blockchain, different?I think blockchain is going to be something – something very important, like spreadsheets important – but the foundation for Web 3.0? Not so sure.JLMwww.themusingsofthebigredca…

    1. LE

      I think blockchain is going to be something – something very importantI am not sure of that by any means myself personally. The reason is anything I think about it is going to be based on what I read (that promoters and detractors say) vs. what I know that others don’t based on my long term knowledge and experience.So for example you could make that statement about Real Estate because you have a history and seat of the pants feel for it in some way. And you have seen patterns. So you can be fairly confident that a pattern interpreted in some way might lead to success because you have seen it before. Or about something that might happen in Austin Texas or even the military.But with crypto what you see is a bunch of people who essentially (and mostly) aren’t involved in something else [1] that are spending a great deal of time getting involved in this. (And people investing spreading bets). So sure that could easily lead to something. But it is not being driven in the same way as the Internet was or even any technology. And when Watson of IBM made his statement about only needing 5 computers for the entire world (or whatever he said) if you had asked him ‘what if the price were X’ he would have revised his estimate. Would have been easy to see the potential.[1] In the 90’s most people involved in the internet were there because they didn’t have golden handcuffs to something more solid. So you typically didn’t find people giving up law firm partnerships or leaving Wall Street banks to gamble on the internet. You found people that weren’t really doing anything (to make money at least) that got involved because they had nothing to lose and in part many of them were young enough to be able to gamble and take a chance.

      1. JLM

        .One of the quandaries of life – what is the value of experience when nobody has any experience? What value your instincts when no experience?JLMwww.themusingsofthebigredca…

        1. LE

          Young people are often dismissive of how much experience matters. The only time I have seen experience not matter is when it allows you to take a foolish gamble that someone with more experience stays away from.I have seen this with doctors fresh out of medical school. They will completely pass off any suggestion that they will get better with actual practice and more experience. It is so funny.Ditto for politicians. It’s kind of laughable when you think that voters think that a person fresh out of ‘a bar tending job’ will somehow be able to get beyond that lack of experience about how things work (when you have to deal with others as you do in politics). I am not saying there aren’t things that person will be able to achieve with a fresh approach. That is not my issue. It’s just the blind optimism over and above what would make sense. So naive. Not even understanding the battle or the history. And not even having any experience in any other similar battles.Who do you want doing the cave rescue if you are in the tunnel trapped by water?Who do you want flying the long haul flight to Singapore in the cockpit?Who do you want doing your surgery?

          1. JLM

            .It was interesting to see how those Thai SEALs planned, rehearsed, and executed the rescue. Very professional. No bailout once they began.A shame that one of them died in the process, but it was a very well planned and executed rescue.I remember the first time I ever built a floating bridge across the Imjin River in Korea and the Rhine River in Germany. It was daunting, but I surrounded myself with guys who had done it before and used to say, “What is your recommendation, sergeant?” a lot of times.After I mastered it, I tried to be equally generous in my assistance to others who were learning how to do it.JLMwww.themusingsofthebigredca…

    2. JamesHRH

      Incremental is the key word.

  16. Frank W. Miller

    A little more general comment. The problem with stuff like this (I see this as one example of Fred throwing some pitch he’s seen in the crypto space against his blog for feedback) is that its not rigorous. There is a nugget of goodness technically in blockchains. However, the work seems to be progressing in the business arena with market hype and lots of wannabes that don’t have any idea what they are actually talking about.When the basic protocols of the Internet were developed, they were done by serious scientists and engineers working in labs and universities with sound technical principles as their guiding bases. I see none of that in this “cryptocurrency” space. What I see is a couple of interesting technical nuggets (the basic combination of technical elements in Bitcoin and the interesting layer on top of that with the language built into the transactions in Ethereum) and then mountains of hype and hope from people that were in many cases lucky to be involved in the steady evolution of the Internet earlier.Where are the Tim Berners Lee, Bill Joy, Vint Cerf, W. Richard Stevens, Doug Comer, and the host of other serious engineers that actually built this stuff in the crypto space? The reason the Internet is successful is because these serious engineers thought about the basics of what they were working on and then published it openly so everyone could understand it. Instead we have an anonymous Satoshi and VC backed companies full of financial people that don’t know what they are really discussing.This is imho the basic flaw in all this stuff. It won’t get any better until some folks in academia and perhaps industry labs get this and decide to go off and develop what these guys want to call this Web 3.

    1. Richard

      Astute remarks. There are 1000s or PHDs and Darpa type research centers that are designing / building things VCs can’t and don’t follow. Many VCS get confused when they believe they are in the field of technology forecasting. They are not. They are in the business of buying low and selling high.

    2. Lawrence Brass

      Very accurate.I think that back then R&D was more institutional and structured. Projects depended heavily on government and academic programs.Opensource and capital availability changed that a bit. R&D can be financed now by private capital represented by VCs, more flexible, less structured and with looser governance in some cases. The latter not always a good thing.

      1. Girish Mehta

        There seems to be an increasing assertion on the net that the strongest technical minds of our time are moving into crypto and blockchain…a massive brain influx of the best technical minds into blockchain.I am clueless about whether this is true.Thoughts ? To what extent do you think this is true ?

        1. Frank W. Miller

          Its a good question. I see some sparks. For example, IBM has embraced blockchain as a basic technology for apps other than cryptocurrency. I myself have over time begun to see that the base blockchain is an interesting implementation vehicle. That didn’t happen overnight, to some degree because of all the bluster around the CC’s, but I do think there is something there. I have to believe that I am not the only one.

        2. Frank W. Miller

          Here’s a great example of what I mean:… Note brief reference to selective endorsement vs proof-of-work. This technical issue is quite important, having the potential to eliminate the horrible power consumption issues with BTC et. Al. but very few know about it.

        3. Lawrence Brass

          Considering the amount of capital crypto and blockchain has available now, I guess it is true. Those that are validated as “blockchain engineers” or “blockchain experts” are surely in the best place to be right now. But then, people here always talk about how excess capital might be bad. Crypto is in that situation now.Want to spiff up your boring project and get some capital? Stick a blockchain to it.On the other hand, blockchain is amazing technology. The promise of automated trust, if honored in an efficient way, will indeed change everything. The potential has been proven. I like @William Mougayar’s balanced view of the industry.My problem with crypto is that I have some doubts about the current implementations’ architectural design. This obsession to “be the OS” of some platforms, to run everything inside the blockchain like Ethereum does I think its wrong because it could led to a siloed environment. When you see the papers and the way they picture the stack it is interesting how they minimize the current network technology it runs on, which is basically the internet. You see slides where the familiar cloud representing the net is now Ethereum. It is a higher level abstraction taking the place of the lower level layer it depends on. Subtle messaging. Unconscious or deliberate, I don’t know.I am not a blockchain expert and have pragmatic views about tech things probably influenced by older distributed database technology. I like to visualize it as a distributed transactional system, a trusted component you can use to build systems instead of a system container.I would love to hear what Leslie Lamport has to say about blockchain technology.

          1. cavepainting

            What they say and What they do are different things. It seems to me that many use Blockchain primarily as a means of tokenizing their project, and hoping that the blockchain pixie dust will give them the capital, talent, and wisdom, to crack the nut.There is no question that decentralized data, computation, and money, are transformational.Those dismissing blockchain projects as toys (or) science projects might almost be as wrong as those who have drunk the kool-aid. As Napoleon said, let us engage and see what happens. There is a long way to go yet. We have to work through the hype and noise and who knows, we might really get to the main things that will change the world.

      2. Zi Wang

        Frank W. Miller – all in all, the industry needs both practitioners (hype-driven engineers, entrepreneurs and investors and academia (hyper-rational scientists, professors).Excerpt from Arvind Narayanan — Princeton associate professor The history described here offers rich (and complementary) lessons for practitioners and academics. Practitioners should be skeptical of claims of revolutionary technology. As shown here, most of the ideas in bitcoin that have generated excitement in the enterprise, such as distributed ledgers and Byzantine agreement, actually date back 20 years or more. Recognize that your problem may not require any breakthroughs—there may be long-forgotten solutions in research papers.Academia seems to have the opposite problem, at least in this instance: a resistance to radical, extrinsic ideas. The bitcoin white paper, despite the pedigree of many of its ideas, was more novel than most academic research. Moreover, Nakamoto did not care for academic peer review and did not fully connect it to its history. As a result, academics essentially ignored bitcoin for several years. Many academic communities informally argued that Bitcoin could not work, based on theoretical models or experiences with past systems, despite the fact it was working in practice.

    3. LE

      The reason the Internet is successful is because these serious engineers thought about the basics of what they were working on and then published itMainly though because the Internet (Arpanet) was government funded because of the cold war threat.You can also thank the Soviets for the reason we have GPS (and probably thousands of other things). Satellites etc.

      1. chasrmartin

        People really need to learn to distinguish between ARPAnet and the Internet. Wildly different protocols, wildly different hardware.

    4. SubstrateUndertow

      A more fruitful metaphor for Web-3/distributive-computing complexities maybe that of “organic-living-systems” a decidedly decentralized trial an error organizing dynamic,Maybe a strictly centralized topdown academic organizing dynamic starts to run out of gas at this level of organic complexity ?

    5. sigmaalgebra

      I essentially agree. For your”a nugget of goodness technically”the research libraries and the weekly guest speaker seminars in graduate schools of research universities are just awash in such that doesn’t get 0.01% of the public or business attention of crypto.So, there is a lot of hype, and apparently so far “goodness technically” has next to nothing to do with the hype.But here is another view: In one word, it’s hype. So, create a lot of hype. There are a lot of miracles in information technology — 14 nm line width, $110 8 core 64 bit processors with 4.0 GHz clocks, 2 TB hard disk drives for $65 or so, Gbps residential Internet — so, mention some obscure technology and turn on the bubble machine.A lot of hype is just 99 44/100% nonsense with the rest water, and a lot of people make a living pushing such hype. There are some gullible “convenient idiots”. One of the examples I can’t forget is Newt Gingrich with two coeds at Amherst:…The media continues to get large audiences and eventually, if only via the Goebbels’s technique of repeating often enough, have some influence. And the media’s standards for what they will publish are about as low as the laws and legal system allow. So, people who want some influence just come out with some hype that the media will publish. That later the hype was shown to be nonsense is little or no problem for the media. E.g., the media was terrible in Jefferson’s day; since then we had Yellow Journalism; now we have the NYT — awash in hype and nonsense. The media loves to publish hype, even total nonsense hype.Such media has stimulated wars and gotten a lot of people killed. So, the media will have no moral objections to pushing crypto hype.So, crypto can be an opportunity for pushing hype. That way, some people can get some attention maybe not yet on prime time on CBS, CNN, MSNBC, NBC, NYT, WaPo, etc, but still some attention.For a VC looking for deal flow and making money, apparently might as well get on the crypto bandwagon and beat the drums and blow the trumpets. There are two cases depending on reality: (1) If crypto is not quite all hype and eventually in reality there is some significant business, then a VC known for being highly interested in crypto will have an advantage in the investments and money to be made. (2) If crypto is all just hype and there is no significant business there, then there still has been some publicity, deal flow, etc. that might have some financial value and likely with little or no downside for the hype that proved worthless.A current, huge example of hype based on technology is AI.Never is a long time: It’d be tough now to show that crypto will never have any business value. So, with low standards for what is worth attention, the crypto hype can continue.But the basic standards of sending money to technology projects remain — “a nugget of goodness technically” is not sufficient for an investment, not sufficient for a university, the NSF, NIH, or US DoD problem sponsors, research departments of corporations, etc.E.g., even in blue sky, ivory tower academics, the usual standards for publication are “new, correct, and significant”. For “significant”, uh, the Higgs boson counts and “a nugget of goodness technically” likely doesn’t. Even in computer science, as a prof explained to me with intense seriousness, a new feature for a programming language doesn’t count, “would be academic suicide”, but a provably fastest possible algorithm for some problem in, say, string search might count.

    6. DegenerateCoder

      The calibre of engineers working on bticoin isn’t too bad. I think once BTC establishes itself as the one and only worthy protocol all the engineers working on BTC now will be known as the future Tim Berners Lee etc.. Maybe this time its Adam Back, Gerg Maxwell, Peter todd, Pieter Wuille etc..

    7. cavepainting

      Even if it is possible to build an end-to-end decentralized app that has decentralized identity, decentralized data, etc. the discovery part -i.e. how a dApp will be found by its users is a major unknown question.On the internet, Google is the primary gatekeeper.On mobile, iOS and Android are the gatekeepers. The devices and OS they built gave them the advantage.In this new world that seeks to abolish gatekeepers and where most blockchain projects are seeking to re-imagine some kind of a network, how will the initial catch-22 be broken? How can dApps get to any kind of liquidity from a usage standpoint ?Approaches like Blockstack app store or Toshi may not be the answer either because they need to be found as well and each imposes its own standards on the user.In my view, the existing status quo of centralized apps will need to interface with the dApp world so that blockchains and crypto can be subsumed into the background and users get benefits without needing to rewire and change behavior.When the tech is still evolving, maybe it is better to focus on getting it right, but we cannot forget that figuring out distribution is often just as hard.

    8. Twain Twain

      No one knows the problems they’re causing in the architectures of the Web — least of all the VCs who may want to “do the right thing” but don’t realize that he male engineers they’re backing DO NOT HAVE THE RIGHT CODE TOOLS (@fredwilson:disqus).For example, Silicon Valley really has no idea the bad and biased data+AI it’s shipping. All because even their most technical people have not really thought through the inheritance of machine logic that we have and whether that’s representative of human language, reasoning, cultures and decision-making.Atlantic, March 2017: “Leibniz’s dream of a universal “concept language,” and the ancient logical system of Aristotle… These axioms weren’t meant to describe how people actually think (that would be the realm of psychology), but how an idealized, perfectly rational person ought to think.“*…Five points:(1.) Leibniz’s binary language is NOT universal. He rotated Chinese hexagrams so that they looked like vertical 0 and 1 — which caused VITAL LOSS OF INFORMATION.(2.) Aristotle’s logic is for a non-existent person. It’s unnatural to assume perfect rationality and so all Machines built on that basis CANNOT reason and understand language like humans.(3.) Even if psychology is used, there’s the WEIRD problem of US psychology models being based on Ivy League students and NOT being representative of general populations. Plus there’s the probability bias & reproducibility problems.(4.) Boole built on top of Leibniz so his work also has that vital information loss.(5.) Shannon builds on top of Boole so also has that vital information loss. Plus electric circuits are NOT how the human brain works. Our neurons work by organic interactions — not simply a series of parallel, binary, linear electro-mechanisms.Net-net, we have the WRONG BASIS of Web stackif the objective is to build human-like intelligence that understands human languages, can do common sense reasoning and support good human decision-making.In terms of Web3 stack, where this seriously matters is:* Transition State Machine* Consensus* Dapp Browser* Query layerIt then affects:* Governance* Messaging* File storage* IdentitySo all of the problems that existed with Web 2 stack are being replicated in Web 3 stack — only this time it’s worse because it means data is IMMUTABLE.Therefore any biases which existed previously stays on the chain and makes the AI go awry.InformationWeek, Oct 2017: “The best example of [bias in data] is the 2008 crash in which the models were trained on a dataset,” said Shervin Khodabandeh, a partner and managing director of Boston Computing Group (BCG) Los Angeles, a management consulting company. “Everything looked good, but the datasets changed and the models were not able to pick that up, [so] the model collapsed and the financial system collapsed.”The technical folks and the VCs really have not thought about unintended consequences.Given what happened with FB and its unintended consequences, it’s vitally important that they do think things through properly this time because, for sure, this Web3 stack has the wrong logic inheritance.

  17. Amar

    I want to take this seriously but there is too much marketing friction/noise in that stack. This reads like decentralization somehow jumped the curve into Gartner style marketing architecture without really earning that right. Stacks (beyond the architectural layers) should be crafted in response to market demands forcing scale.As it stands, this seems to be driven by VC demands that wish it to become mainstream now rather than later. I am not being cynical and I think that is the job of a forward looking VC. But, it would gain credibility if it was use case driven.

  18. JaredMermey

    Is there enough incentive to ensure all layers of the stack are built? If token accumulation/appreciation is main incentive, will “market” guesstimate which protocol might win and speculate the associated token before the dApp layer comes leaving little room for realized value for the application layer builders (let alone middle layers builders)? Is liquidity and transparency of the token market both a gift and curse?Perhaps these questions seem too broad to the point they are platitudinal and show no real insight, but it feels like for this to become Web 3.0 some equilibrium of value distribution has to be formed. I am not looking nearly as hard as others who play here everyday, but from laymen’s/outsider view do not see it yet (or a path to it yet).

    1. kidmercury

      incentives is one of the main reasons i believe the blockchain will end up creating all these elements and delivering them for dapp developers, essentially creating the architectural blueprint for dapp developers rather than allowing/requiring them to create it themselves. the blockchain operator controls the reserve currency of the whole ecosystem, and thus can dictate the flow of capital across components.

  19. Seth Ona

    What is planetary consensus? (just below of the state machine)

  20. Amar

    Btw, I just saw this commercial on TV. I am in London today and this ad played during the football half-time … It makes no.freaking.sense…This is a classic example of peak marketing hype with ABSOLUTELY nothing to show. Why would you run this advertisement during the football world cup semis when it is 100% vaporware? To sell ico’s of course #SMHIt is this sort of behavior that makes it hard to take most of the blockchain hype seriously.

  21. sachmo

    Imagine an app that notifies you when food trucks are within a few blocks of your location built on the blockchain in a manner that protects your privacy.On Web 2.0 I’d argue that it is still the case that a talented person can still hack something together by themselves or with very little support and no outside funding — but it’s getting harder to do that.If your model of Web 3.0 is correct, it hurts my head to think about how many layers of programming might be required to release an app or design piece of hardware connected to the internet from the ground up.I wonder if an individual will still be able to do that in the future, or if they will be forced to sit on top of Dlayers of protocol and pay hefty licensing fees.

  22. Pascal Aschwanden

    I would like for Web 3.0 to be every bit the success that you are alluding too. As developers, entrepreneurs, as VCs, and anyone else who wishes to see progress, we all want there to be a Web 3.0. No one likes a world trapped in walled gardens. And, we can all imagine the enormous benefits of a world where people truely own their own data. But, can users? How do we know it’s not a pipe dream?At end of the day, B2C Users don’t care about Web 3.0 or any other technology stack. They care about a solution. Is there any evidence that users will switch from existing social networks for the sake of data ownership? I know there are many other benefits but I don’t think users will switch services for some grand vision of the future (as much as we wish it were true)As, you said it in your other post, I hope it’s just a matter of finding the right delivery mechanism, the wrapper you spoke of.

  23. Jordan Taylor

    @fredwilson:disqus You say ‘you are more sure about this than anything else right now’ That’s a powerful statement. Is there anything you or your colleagues have written in the past you can point me to that rigorously makes the case as to WHY a decentralized linkedin or facebook would actually be better than what exists today? I’m not dismissing your assertion, but would really like to see more rationale behind it. Thanks!