Video Of The Week: Brad In Berlin

Our partner Brad has had an interest in Berlin’s growing tech ecosystem since the earliest days of USV. One of our first investors in USV is a good friend of Brad’s from Berlin. So he has been a big supporter of what is going on in Berlin and goes there frequently, as we all do. He has developed a close relationship with Blue Yard, a new VC firm in Berlin, and he and Blue Yard’s co-founder Ciaran O’Leary did a talk there last month. It’s really good (and not too long at 24 mins).

#blockchain#VC & Technology

Comments (Archived):

  1. Twain Twain

    Wrt authenticity & integrity … Network effects + AI AMPLIFY BIASES, including sexism (see MIT Technology Review article on MS Research into Google Word2Vec).Blockchain being a network that, potentially, REPEATS THE SAME MISTAKES AS INCUMBENTS should not pass us by.

    1. Twain Twain

      Some core problems in database infrastructure and data structures need solving so that SYSTEMS are FAIR, REPRESENTATIVE & SERVE ALL OF US.This is especially important because AI makes the risks of network effects amplifying and propagating sexism, racism, etc.(via Deep Learning layers) even higher.That’s why instead of “making money hand over fist” in the Blockchain sector, I’ve been inventing to de-risk network effects in Blockchain & the amiss AI frameworks provided by the likes of Google, Stanford and Microsoft itself.Yes, MS Research is right to call out Google+Stanford Word2Vec as being “BLATANTLY SEXIST”. Nonetheless, Microsoft should look to its Tay bot to see that their AI’s biased too (sexist, racist etc.) and the networks effects of spreading that bias fast also affects them as much as Google.

      1. Twain Twain

        Incumbents & Blockchain upstarts use similar network, data and probability+statistics structures.Decentralization and encryption does not solve that the structure & methods provided by probability default to amplifying biases.Investors need to reflect on whether they’re unknowingly rubber-stamping and investing in systemic bias (sexism, racism, etc.).Well, of course … there’s a solution …Kick the butts of Descartes logic + Bayesian probability (basis of network biases) with Da Vinci.:*).

    2. LE

      There is truth to the doctor nurse thing. It’s been well known for a long time actually. Is it really a problem if AI recognizes and uses stereotypes if the stereotype is correct?

      1. Twain Twain

        It is a genuine problem because those stereotypes and biased frameworks of Descartes and Bayes’ would only be “valid” for the time BEFORE women entered education and workforce in such increasing proportions. They’re anachronisms from C17th and C18th when women were mostly excluded from high school (much less university), not C21st.Frankly, those stereotypes and frameworks are no longer “fit for purpose” and representative of us.In 2015, in the US, the ratio of female:male medicine graduates was 0.9 : 1 which means there’s almost an equal chance of y = female. So that’s a fundamental example of how wrong and wide of the mark Google’s Word2Vec framework of “father : doctor :: mother : x” and it will say x = nurse is, GIVEN THE DATA.Meanwhile, in the UK from 2007-2012, the number of female doctors under the age of 30 had increased by 18 per cent, while the number of males decreased by 1 per cent. That macro trend means that by 2017 … FEMALE = DOCTOR is more likely than male = doctor in the UK.@fredwilson:disqus is always, rightly, saying we need to look at the data.So … given the data … we need to ask:(1.) Is the right type of data going into the systems?(2.) If Descartes and Bayes are fundamentally biased and unrepresentative of us, how can we invent and solve this problem so the systems are fairer and more representative of our modern society and for future generations where girls may grow up with a female President and female CEOs of Fortune 100 companies and female Heads of the IMF etc (not the bygone society during which Descartes and Bayes created their frameworks)?

        1. LE

          I don’t know anything about this but I will assume baked in is the ability to change with the times. Isn’t this the case? Also geographically and based on other inputs?By the way I still think that the probability of a wife being a nurse if the husband is a doctor is greater than chance despite the fact that more women are in that profession than 100 years ago.The reason is young doctors are exposed to more female nurses fresh out of medical school than they are female doctors. Also the female nurses tend to be more subservient than the female doctors. As a result the men will be more likely (along with perhaps looks) to hook up with a female nurse especially because of the power dynamics in the work environment. And many women (not all of course) become nurses so they can “marry a doctor”. Terrible thing to say? It’s somewhat true. Like saying “many men become doctors for the money”.

  2. awaldstein

    Really terrific. Possibly the best I’ve listened to.If I could ask a question from the audience I would ask:”You started with the statement that your bet is that the future will be ‘less about market power and more about user experience’. Can you clarify what you mean by user experience in that content as you never close the loop on this term throughout the talk and it is key to me?:Fred or Brad–would really appreciate this answered if you care to spend the time.Thanks!

    1. fredwilson

      I think what Brad means when he says that is businesses will have to compete for our loyalty every day because they won’t have market power that locks us in. And in that world we will have lots of UIs to choose from and we will use the one we want. Kind of like email clients and web browsers

      1. Guy Lepage

        To add to this a bit. At first there will be quite a few UI’s but as these systems grow, you’ll see dominant UI’s that will emerge. Not only do these systems generate major innovation on the protocol layer but will in turn create a new competitive UX/UI layer like we’ve probably never seen before. I say this because the current state of UX/UI is at a sophisticated level and we will now have to rethink these this layer in order to be dominant over the long term. IMO

        1. Twain Twain

          And who sets the standards in UI/UX? Apple.And look at the master-stroke they just pulled in acquiring Turi (formerly Dato), an AI startup.I was at Turi’s conference in July where they showed how their system is 3x FASTER than Tensorflow.

          1. Guy Lepage

            I personally believe that it will be a combination of standards some at the protocol layer setting high level requirements/restrictions and some standards will be set by the early dominant apps. I believe there is a social and design responsibility for the protocols to think hard about what these new systems will bring to society and to make the new web a better, richer, and more beautiful experience than it’s predecessor.

          2. Twain Twain

            Well, in AI era, there’ll have to be standards at the protocol layer setting high-level requirements / restrictions that HALT the AI from launching nuclear attacks.How this will happen is open for debate since a combination of Turing completeness and Gödel’s incompleteness theorem remain computationally unsolved (note, that doesn’t mean unsolvable).There’ll be new design standards that hybrid Web and mobile such that the user experience is the same (no need for separate mobile app).There’ll be new data integration standards between data collected via Web and data collected via mobile. Currently, the whole data munging process is costly and inefficient.Yup, Brad wants innovation that isn’t necessarily coming from the incumbents (and their locked-in data and network power effects) and he can rest assure there’s still plenty of innovation ahead for all of us.

          3. Rick Mason

            But will Apple make it as freely available as Google has done with TensorFlow? Along with the educational materials so the average developer can make use of it?If Apple had opened up Siri they could have had someone seize the market Alexa currently owns.Because as amazing as Google or Apple’s engineers might be it is always that team in a Detroit garage that will succeed in doing something no one else has even imagined.

          4. Twain Twain

            Apple has made SIRI available to developers:*…ALL of them, including Amazon Echo, are still missing that “something they haven’t even imagined” though.Why? They default to logic.

          5. Rick Mason

            Sure Apple has made Siri available to developers in June of 2016, but only after the success of Alexa shamed them in to doing so. Alexa welcomed developers long ago to the party and there are hundreds of skills available for it. Siri has a lot of catching up to do.If they wait until 2018 to make Turi openly available it will be too late. The march of progress does not wait for any company.

        2. Twain Twain

          Totally agree with your comment: “new competitive UX/UI layer like we’ve probably never seen before…rethink these this layer in order to be dominant over the long term.”Indeedy and 99% of engineers have no clue how to design or code this UX/UI layer because all they did was repurpose Web 1.0 format interactions for smaller screen size of Web 2.0 mobile.SV hype machine wrongly assumes we can go straight to voice or chatbot.But there’s another vital UX/UI layer to nail before the voice AI and text chatbot can understand what we mean by what we input.

          1. Lawrence Brass

            For me, mobile most important contribution (aside mobility, of course) was the touch interface. Size in mobile is an implicit restriction. Still a hands on experience. Future is hands off, voice recognition beyond speech-to-text, non touch gesture recognition, eye tracking. Sensors.As you always teach us, the human love needed to process this sensed data is crucial. There is no Artificial Love yet.

          2. Twain Twain

            Thanks, Lawrence — exactly and as shown in this picture I took at Comic Con SV.The “fathers of AI” are running around trying to teach the AI our natural language, in the total absence of “mothers of AI”.That’s Google Brain’s Word2Vec team (including Jeff Dean who heads Tensorflow), Stanford’s GloVe team which includes Stanford’s Chris Manning, MS’s team that built Tay bot, IBM Watson whose own inventor called it “akin to a human autistic savant,” Elon Musk’s OpenAI team whose Research Director co-invented Word2Vec and the Allen Institute.

          3. Lawrence Brass

            Fathers of AI should take note that “call Mom” is more prevalent than “call Dad”, as it was commented here months ago. Was a google trend data comparison, if I recall. Nothing more powerful than mothers’ love.AI will be autistic until some artificial love is infused into it, surely coded by a woman.

          4. Twain Twain

            President Obama, 4 August 2016: “Michelle and I have raised our daughters to speak up when they see a double standard or feel unfairly judged based on their gender or race—or when they notice that happening to someone else. It’s important for them to see role models out in the world who climb to the highest levels of whatever field they choose. And yes, it’s important that their dad is a feminist, because now that’s what they expect of all men.You see the subtle and not-so-subtle social cues transmitted through culture.”*…@fredwilson:disqus @wmoug:disqus @MsPseudolus:disqus – It’s not simply the maths I’m solving, it’s also the culture.It’s a culture in which an unrepresentative clique of a dozen Professors and PhDs of AI are (unwittingly) top-down shipping, cultivating and network-effect transmitting sexism and racism in the form of Word2Vec and MS Tay bot.

          5. Lawrence Brass

            My pleasure, Twain.

          6. Twain Twain

            Interestingly, see Turing and Babbage’s quotes. Investors never expect it to be women who solve what guys can’t … LOL.

          7. LE

            sometimes it’s the very people who no-one imagines anything of who do the things that no-one can imagineThat’s really just the law of large numbers at work.There are a ‘jillion people out there that nobody knows about, or has heard of, doing something. Quite a bit more than people who we know about or have heard of. And if enough people are trying enough things something will work and they will then be known as will what they have discovered or had success in. [1] Unlike in sports there is really no farm team or scouts to discover talent early on, although one exception that I can think of is VC’s and angel investors who act as a filter. (And of course top Universities). But mostly perhaps there are people like Ray Kroc the multimixer salesman who happened to stumble upon the McDonald brothers one day and got an inspiration.[1] Enough s**t at the fan and something will stick.

          8. Twain Twain

            Modern-day Ada Lovelace(s) framework and encode the representative data structures of “Artificial Love” into the machines.Unfortunately, men — spanning Turing and John Von Neumann to Noam Chomsky to Marvin Minsky to Google+Stanford+MS+IBMWatson+Facebook+Intel+Amazon+++ —have had since the 1950s to solve it, and FAILED to do so.So she does indeed framework, design and CODE IT.Feb 2011: She mentioned to Ray Kurzweil she’s working on making AI more emotion-able. He said, “That’s the most cutting-edge area of AI.”June 2014: She saw a Kurzweil TED Talk in which he argued that the increase in our brain’s mathematical ability enabled us to invent arts and sciences, and she knows this is not exactly right.After all, our cave paintings & Chinese calligraphy PRE-DATES Sumerian and Babylonian maths.Oct 2012: She asked Amit Singhal, “Does and will Google’s Star Trek engine have a heart?”Feb 2015: She asked Greg Corrado, one of Word2Vec’s inventors, “How does Word2Vec deal with the subjunctive tense” (which involves emotion expressions over continuous time).13 July 2016: She asked Jeff Dean of Google Tensorflow if Word2Vec is the right structure for understanding Natural Language.All of them were amused and reacted to her as if she was either a journalist (so didn’t know the maths involved) or just plain crazy to question Word2Vec.21 July 2016: MS Research publishes paper showing Google Word2Vec is biased (sexist, racist etc) and that bias is being amplified by network effects with Deep Learning.@fredwilson:disqus – I realized this from the start: Google & the incumbents’ are disciples of Descartes and Bayes.Meanwhile, I’m a disciple of Da Vinci and the I Ching (Yin+Yang in coherency).This January 2016, I was in the room when Chris Manning of Stanford who led GloVe project (which has similar biases as Word2Vec) said: “Higher level language is of a DIFFERENT NATURE to lower-level pattern recognition (aka by probabilistic+statistical methods).”Well, Da Vinci is of a DIFFERENT NATURE to Descartes.Da Vinci is heART and science. Descartes forced us to separate mind from body from emotions to arrive at “Rational Logic”.So NO AI or economic system built with Descartes as a basis can ever represent us. That type of AI is actually pre-biased against us.Only AI and economic systems built on Da Vinci’s basis will represent us.No inventor or founder has been on a mission like mine, :*).

        3. sigmaalgebra

          A Stagnation ManifestoSorry: IMHO for at least another 20 years, nearly all business significant UIs will be based heavily on the now very practiced controls in HTML — links, single line text boxes, multi-line text boxes, check boxes, radio buttons, and relatively simple extensions from JavaScript and for touch screens and mobile.Why? (A) That is what exists, is known to work very well end to end from the UI designers to the users, and is solid in nearly all respects. (B) That’s what 2+ billion people already understand very well. (C) That’s both what and about all the software developers know. (D) Newer approaches to UI will be for special purposes for a tiny fraction of the users/usage.Or, in simpler terms, for a UI over the Internet, HTTP and HTML are by now just too darned well designed and established.Besides: HTTP and HTML are fine for delivering much more utility for the users. That is, for what is really important for the future, we don’t need to replace HTTP and HTML.By analogy, HTTP and HTML are like the bottle, not the wine. What’s of nearly all the importance is the wine. Yes, the UI/UX has to be good enough, but it’s never more than the bottle and never the wine.Replacing HTTP and HTML is about like trying to replace, in cars, a 15 gallon tank of gasoline, a V6-V8 engine, a liquid torque converter, an automatic transmission, and rear wheel drive with a center differential. Replace carburetors and distributor points with computer controlled fuel and ignition? Yup, great. Replace the rest I mentioned? Nope — it’s just too darned good, too well refined, too well understood, too well accepted.People have been working on UI since teletype. There was more with IBM’s very carefully designed 3270 series, a copy of much of that for dumb terminals, graphical user interfaces, e.g., Microsoft Word, general purpose GUIs, e.g., X-Windows, lots of word processing languages, e.g., Runoff, Script, Scribe, and TeX, various touch screens, stylus interfaces, and then the Web, HTTP, HTML, JavaScript, the refinements to the present, and the refinements for small mobile devices.But for getting rid of the keyboard, UI from talking to the computer, having the computer pay attention to eyeballs, facial expressions, hand gestures — no thanks.Again, sure, for some special purposes, e.g., games, engineering design, architectural drawings, creating artistic images, image editing, video creation and editing, ad copy creation, music composition, and more, sure, maybe have something special.Net, for 2+ billion people using the Internet, IMHO it stands to be essentially just HTTP and HTML, for at least 20 more years. Sorry ’bout that.

          1. Michael Elling

            Mostly agree; despite the infinite range of demand, current business models dictate satisfying the largest demand sets around the mean. Variability is eschewed; even if the latter provides the optimal user experience. We’re still in a kilo and low-mega world even though the technology is giga.What would change that (providing greater divisibility) is: a) balanced settlement models affording “centralized procurement” of edge access (and user experience), or b) a rapid move away from mostly text to video and audio user experience model. I would bet on the former happening at scale before the latter over the next decade. Just look at how pervasive messaging apps are and the challenges those platforms are having with 1st gen chatbots.”Centralized procurement” is counter-intuitive to people’s thinking of edge variability, but it will actually result in everyone’s unique demand curve being realized; thereby clearing marginal cost and marginal demand rapidly and solving the problems of rapid capital depreciation. And yes, this is a violation of the contrived notion of “net neutrality”.

        4. Michael Elling

          UI/UX is limiting. Look at the complete user experience. Likely no two smartphone home screens on the planet are alike. Start with that assumption. Now build an experience model (all the way down to access and transport) that fully satisfies all that varied demand. That’s where we are heading.

      2. jason wright

        Is this the coming of the “constellation” pattern you wrote about a couple of years ago?

      3. awaldstein

        Got it.Spent the last month reading and listening to just about everything on this topic as it is touching my world.Brad is saying (to me at least) that we (USV) are making a bet that X could happen. That there is a ton that we don’t know but the patterns point in a direction of possibility and that will touch trust, empowerment and challenge the status quo of market forces.This is actually a bold huge leap that i can begin to wrap my head around.

        1. sigmaalgebra

          That there is a ton that we don’t know but the patterns point in a direction of possibility and that will touch trust, empowerment and challenge the status quo of market forces.IMHO, all that is so vague and difficult to see that no one could be very sure it exists at all and even if it does exist has such a tenuous connection with business it is a long way from business success.E.g., need a lot more than just “possibility”. And “touch trust, empowerment and challenge the status quo of market forces” IMHO as a foundation for a business is about as solid as soaked tissue paper.In particular, similarly for the Blue YardWe are a thesis driven firm that backs founders decentralizing markets, empowering users and liberating data.Are these three really solid and important, nearly necessary, ways greatly to please, thrill, the users/customers and build successful companies or mostly just some social objectives?Or (A) as a user/customer buying from a market, given that I like what I’m getting, why do I care very much if a market is “decentralizing” or not? (B) For “empowering users”, it would appear that computing and the Internet have users much more empowered than ever before. (C) For “liberating data”, any user with access to the Internet is awash in liberated data.

          1. jason wright

            Would you regard your personal identity data, if legally owned by say Zuckerbook, as liberated?A multitude of personal data permissions rather than the data dictatorship of ZB is democratic data liberation.

          2. sigmaalgebra

            If there is an issue about what Facebook, Google, Amazon, the many companies placing cookies, etc. have on people, then, sure, pass some new laws. But to me that’s a political/legal issue, not a technology, VC, or business issue.E.g., for liberating data, we already have laws about credit scores and the Freedom of Information Act. If we need more such, then use the political system. For the technology part, Firefox has ways to throttle cookies; there are laws about giving your e-mail address to spamers and also the spammers.But, realize, if you shop at a retail store that uses a bar code scanner and pay with a credit card or check or show a discount card of some kind, then the store knows item by item everything you bought, and that’s been true for a long time.

      4. LE

        that is businesses will have to compete for our loyalty every day because they won’t have market power that locks us inBusinesses needing to wake up and compete is always good. Businesses that don’t get fat and lazy and eventually are replace by a more motivated competitor.Look at the movie (or broadcast TV) business. They have to bust their ass year after year to come up with hits and success. (Sports is similar.) As a result they can never rest on their laurels and take anything for granted (no cash cows) since they don’t have a lock on anything they are only as good as their last hit. That lack of defensibility I feel is what has allowed many of the players that were around in the early parts to still be around today in one form or another.And although autos lost their footing when the Japanese came and had their bankruptcies they also are a business where you have to turn out new models year after year or you will loose out to competition and most importantly you know it on a quarter by quarter basis. (And they are all still around mostly..)

        1. jason wright

          Their blockbuster productions are a barrier to entry strategy. So very expensive.

      5. Michael Elling

        Infinitely divisible demand. To achieve that we need centralized procurement via settlement systems that don’t exist yet; all the way down to the physical (access and transport) layers.

      6. Matt A. Myers

        This is through governance, of course – the final layer of competitiveness, the user voting with their time; governance is leading a community of people, educating them on what you are about and then maintaining integrity and being transparent about the processes for maintaining said integrity; Elon Musk seems to do a good job of this – a great role model and inspiration for us all.I touched on governance a little in a comment last week at Continuations, and I plan to write a more full blog post based on the comment, prompted by Albert;

  3. jason wright

    When a promising blockchain business forms as a foundation and not an LLC the traditional equity stake investment model is not an option for a VC. Buying coins is an option, but the concentration of coin ownership in the hands of a few would undermine the growth and success of the blockchain. It’s a dilemma it seems.

    1. awaldstein

      any discussion of business formation without addressing the issues of taxes honestly to me feels like mental exercise.none of this will happen until the legal issues are addressed. both will but taxes are the basis of the world and a decentralized system will not grow without smacking into it.

      1. jason wright


        1. awaldstein

          Not understanding Jason.Are you seriously suggesting that as the world and commerce becomes decentralized (if it does) the core business model will become stateless and move to the Cayman’s?

          1. jason wright

            The tax gathering nation state is an artificial evolution of the tax gathering city state. Both are functions of geography. The central bank fiat system is an instrument of the ns system, and there’s nothing natural about it. We simply don’t know any better. Now I see a basis for change.

          2. awaldstein

            Then how do potholes get fixed and police to keep me safe?

  4. pointsnfigures

    one thing to pay attention to is the big platform internet companies. watch Facebook, Google, etc-and their level of lobbying. Big players will lobby for regulations to stop competition. It might make it harder to have a bunch of dispersed companies.

  5. Ciarán O'Leary

    What sets Brad apart for me is his ability to so incredibly clearly articulate a mission for both entrepreneurs and the VC backing them; what things should exist to make markets and societies better – and then to brake that down in to sequences, technology stacks, etc. It is going to be a lot of fun to see if the new tool kits and protocols we have around decentralization can unleash their potential. I sure hope so.

  6. LE

    people have been able to raise a lot of money at a relatively high valuation and it’s very very hard to turn that down as an entrepreneur but what I think very few entrepreneurs realize is the trap that they were creating for themselves if you raise that money at that valuation you have invited in a set of partners that have an expectation of a return as a multiple of that valuation and that means that somehow someday you have to generate that much value. This reminds me of a lesson that I learned as a kid in my Dad’s wholesale import business (the 70’s). During that rough economic time I remember him laying people off. And that the people that he kept were the ones that provided value and hadn’t pushed for the last dollar in wage negotiations “because at that hourly wage I can always find something for them to do”. The higher paid people, those who negotiated and demanded the highest wages, were the ones that lost their job.

  7. William Mougayar

    More decentralization is part of our future, and Brad has articulated his thoughts very well.How we will get there is going to be via some trials and tribulations.

  8. Vendita Auto

    Honest intuitive replies on the Q&A’s A joy to note the the personal knowledge base

  9. LE

    I don’t see any arrogance at all in trying to shape a business that a VC has invested in in their own image or to their own needs. Stop hating the game as it is played. And stop whining. Nobody is forcing anyone to take money. If you take money you take the risks that someone else will have an opinion and control. If you don’t want to take money then you can stay at a small scale and fight and commiserate with the rest of the cry babies.

  10. Kirsten Lambertsen

    Enjoyed this one a lot. Not the same ol’ same ol’ (aka ‘is there a bubble?’). Even the questions from the crowd were different than the usual. It’s interesting to think about how reiterating “I’m a capitalist” might strike a different chord for audiences outside the U.S. than inside.Did the event organizers make the panelists sit on gas cans so that they’d be too uncomfortable to go over their allotted time? 😉

  11. LE

    I don’t get the investor love and confidence in Google and Facebook. If they were so confident of their future they wouldn’t be investing in moonshots. Their chance of succeeding at those things is by no means even close to certain. To me FB is a huge long term risk especially and importantly as their workforce ages. And I don’t know history wise of any company that has been as dominant as either of those that has successfully continued to be dominant by taking the types of risks or making the types of bets (essentially investment bets) that they have been doing. This is really a unique and new business model. I mean I get how they are scared ‘it’s not going to happen to us’ and why they are trying to find a new path or paths. I just think investment wise (stocks) it’s not anywhere near a long term hold.

  12. kenberger

    I attended this Blueyard-produced event and it was fantastic. An extraordinary group of people gathered to discuss some super forward-thinking topics (Decentralized & encrypted was the theme). Blueyard is clearly very much aligned with the USV book of Big Ideas that are revolutionary as well as evolutionary.@ciaranoleary:disqus ‘s partner Jason also interviewed Snowden via video. Fascinating content and I sure hope these guys do more.(man I travel too much, and keep jumping in to these comments a day late!)

  13. jon blondyn

    wow, really enjoyed listening to this. (did so twice). decentralized, emergent, bottom-up innovation. really great to hear the vision. thanks so much for posting

  14. Matt A. Myers

    “… what’s the new source of market power that’s going to catch us by surprise, and can we do anything to prevent that in the way that we design these platforms.”This hit me – a fantastic holistic view of the things to come.I feel the market power that should be obvious, however likely will be a surprise, is that the governance itself of these decentralized platforms will dictate which one(s) become the dominant players; I understand that is the likely the goal, to make the mobility of people online so fluid that swarms (AVC-portfolio company pun unintended) of people or groups can quickly utilize a new platform that better suites their needs – either a decision based on utility wise or beliefs.Maybe we should be already moving forward with making the foundation of the internet, the infrastructure, have this same kind of fluidity? The physical lines that connect our devices, as well as the domain system – e.g. if you don’t like that GoDaddy’s founder hunts animals for sport, then you can move all of your services at the click of a button somewhere else.One thing I can foresee is that the potential volatility could make the actual valuing and sale of a business very difficult, especially if the value and the user base/network on the platform exists due to trusting the governance of one or a few individuals – and if they’re no longer in control and if the principles used for decision making, on every level, have any hint that they’ve lost their integrity then the ecosystem should collapse much more quickly than the historical collapse of dominant platforms (MySpace->Facebook etc).The final importance after governance is taken care of is the brand name and how easily it can resonate with people; intellectual property rights themselves very clearly add friction to rapid innovation – perhaps a good coping mechanism until we have a proper and safe net in place.A radical idea: what happens if I was able to curate my own structure (UX/UI/monetization wise) of Facebook that users could choose to use (“Matt’s version of Facebook”)? That way users are staying on Facebook, however they aren’t at the mercy of how any individual decides what is “best for them.” And not wanting to pollute the clarity of the thought experiment with business modelling, however would it be possible (or even desirable) to have a revenue stream associated with providing the brand-originator as a means of compensation? Or perhaps all of this customization must be internally motivated – which would be difficult for most existing platforms to implement due to fear or simply not having those resources available; those resources may quickly become available if we had a real free labour market via guaranteed basic income.I should probably tie this into my next blog post.