5 ms·

>It's the trend of turning software into services. [...] But it also makes you no longer in control of your data, and the code that runs on it. The availability

by jasode 9y ago

>It's the trend of turning software into services. [...] But it also makes you no longer in control of your data, and the code that runs on it. The availability of your work becomes completely dependent on business decisions of third parties, which can - and frequently do - disappear suddenly. Here's a thought experiment I'd like people to ponder: Suppose we had this decentralized hardware utopia where every homeowner had his own web server node ... such as a "FreedomBox"[1][2] or a hypothetical ISP smart router with builtin 16GB RAM and 10 terabytes of disk to hold Sandstorm[3] or whatever self-hosting app stack you can think of. Even with those favorable conditions, I'd argue we'd still evolve towards Youtube centralization. I'd encourage people to think about why that counterintuitive result eventually happens. (As a hint, Youtube does things that localized technology installed in the home doesn't solve and can't solve.) >If technology is meant to be [...] empowering to individuals, then turning everything into a service is an exact opposite of that. To further explore if eventual mass migration from home self-hosted videos to centralized Youtube is inevitable, we have to be open-minded to the idea that thousands of people will conclude that "putting my video on Youtube instead of hosting it on my own server is what empowers me." Why would people think that opposite thought? [1] https://freedombox.org/ https://freedombox.org/ [2] https://www.nytimes.com/2011/02/16/nyregion/16about.html https://www.nytimes.com/2011/02/16/nyregion/16about.html [3] https://sandstorm.io https://sandstorm.io

mac01021 9y ago
> Youtube does things that localized technology installed in the home doesn't solve and can't solve. Are you talking about search and content discovery? Are you talking about replicating the most popular videos over many servers so that the system can efficiently serve all the demand to watch those videos? Maybe existing, fully-distributed products don't solve these problems. But it isn't obvious to me that they can't.
- matte_black 9y ago
  He is talking about monetization. It’s not even that counterintuitive with some capitalist forethought.
  - TACIXAT 9y ago
    Monetization is easily possible, we just need a reliable way to transfer small amounts of money electronically at a low fee. Right now the large credit card players have a stranglehold on digital transactions, we just need something new that makes it not awful to pay someone a few cents online.
    - kristianc 9y ago
      Consider that the people watching the videos have repeatedly shown in no uncertain terms (even when given the option and it is made really really easy for them), that they don't want to pay.
      - TACIXAT 9y ago
        Really? I absolutely will not pay 8$ a month to read the occasional NYT article. I will not pay 4$ per episode of TV shows I watch (I have tried this multiple times and it sucks). I would absolutely pay 5 to 25 cents to read an article. I would absolutely pay 1$ per episode. Unfortunately, no one is offering their content at these realistic price points. Nor could they with credit card processors taking 30 cents per transaction. We need a new model that makes costs realistic. I bet a lot of people would pay 1 dollar a month for social media to avoid ads and tracking.
        twblalock 9y ago
        Facebook had revenue of about $40 billion in 2017. Facebook has billions of users, so if every user gave Facebook a dollar a month, that could add up to about as much as they made last year. (I don't think every user would do that, especially the users who live in developing countries and can't afford to spend a dollar, but let's use the best case for the sake of argument.) What happens when a startup decides to take on Facebook by offering a social network that focuses on protecting user privacy? We would want that startup to succeed, but how could it ever get there by charging a dollar a month? It would start with a small number of users, so the total dollars per month would be small. If Facebook was only charging a dollar, this startup couldn't charge more than that -- a social network that has fewer of your friends on it than Facebook and costs more to use is not a very attractive. In other words, I think the subscription model would make it difficult for social media startups to succeed.
wernercd 9y ago
One thing I've heard harped on recently is how common the Pareto type distributions are. "80% of the work is done by 20% of the workers". https://en.wikipedia.org/wiki/Pareto_principle https://en.wikipedia.org/wiki/Pareto_principle Where this comes into play with regards to your thought experiment... is how the larger the playing field, the fewer the large players are - they tend to get "concentrated". Plenty of Facebook alternatives exist... but why go to them? No one uses them. Large chunks of friends/family/acquaintances aren't "there" - they are on Facebook. Plenty of authors other than Steven King exist... but given a few spaces on shelves at kiosks, who's going to get those spots? You can guarantee King will get one of those few spots... The majority of activity is going to condense down until a few spots. Why? Because that is where everyone else is. That's where the skilled "creators" are. The question becomes... how do you make sure those spots are fair, open, balanced, secure, etc. Data in the hands of a few companies? What happens when they all lean left - politically? What happens when one of them gets hacked? When a "foreign actor" learns how to game that system? I've thought for the last decade that things will continue to get "worse" before they get "better". How much worse? How can we make it better? Million dollar questions.
- haZard_OS 9y ago
  I may be accused of nitpicking here but the Pareto Principle is, although related, distinct from the Pareto family of distributions. I don't intend to derail the conversation so, rather than typing out an explanation of obscene length, I will refer all interested parties to the following: Pareto Principle https://betterexplained.com/articles/understanding-the-pareto-principle-the-8020-rule/ https://betterexplained.com/articles/understanding-the-paret... Basics of Pareto Distributions https://actuarialmodelingtopics.wordpress.com/2017/05/21/pareto-distribution/ https://actuarialmodelingtopics.wordpress.com/2017/05/21/par... https://subversion.american.edu/aisaac/notes/pareto-distribution.pdf https://subversion.american.edu/aisaac/notes/pareto-distribu... Example of some limitations in the application of Pareto Distributions https://arxiv.org/pdf/physics/0504217.pdf https://arxiv.org/pdf/physics/0504217.pdf

by jasode 9y ago

mac01021 9y ago

> Youtube does things that localized technology installed in the home doesn't solve and can't solve. Are you talking about search and content discovery? Are you talking about replicating the most popular videos over many servers so that the system can efficiently serve all the demand to watch those videos? Maybe existing, fully-distributed products don't solve these problems. But it isn't obvious to me that they can't.

matte_black 9y ago
He is talking about monetization. It’s not even that counterintuitive with some capitalist forethought.
- TACIXAT 9y ago
  Monetization is easily possible, we just need a reliable way to transfer small amounts of money electronically at a low fee. Right now the large credit card players have a stranglehold on digital transactions, we just need something new that makes it not awful to pay someone a few cents online.
  - kristianc 9y ago
    Consider that the people watching the videos have repeatedly shown in no uncertain terms (even when given the option and it is made really really easy for them), that they don't want to pay.
    - TACIXAT 9y ago
      Really? I absolutely will not pay 8$ a month to read the occasional NYT article. I will not pay 4$ per episode of TV shows I watch (I have tried this multiple times and it sucks). I would absolutely pay 5 to 25 cents to read an article. I would absolutely pay 1$ per episode. Unfortunately, no one is offering their content at these realistic price points. Nor could they with credit card processors taking 30 cents per transaction. We need a new model that makes costs realistic. I bet a lot of people would pay 1 dollar a month for social media to avoid ads and tracking.
      - twblalock 9y ago
        Facebook had revenue of about $40 billion in 2017. Facebook has billions of users, so if every user gave Facebook a dollar a month, that could add up to about as much as they made last year. (I don't think every user would do that, especially the users who live in developing countries and can't afford to spend a dollar, but let's use the best case for the sake of argument.) What happens when a startup decides to take on Facebook by offering a social network that focuses on protecting user privacy? We would want that startup to succeed, but how could it ever get there by charging a dollar a month? It would start with a small number of users, so the total dollars per month would be small. If Facebook was only charging a dollar, this startup couldn't charge more than that -- a social network that has fewer of your friends on it than Facebook and costs more to use is not a very attractive. In other words, I think the subscription model would make it difficult for social media startups to succeed.

wernercd 9y ago

One thing I've heard harped on recently is how common the Pareto type distributions are. "80% of the work is done by 20% of the workers". https://en.wikipedia.org/wiki/Pareto_principle https://en.wikipedia.org/wiki/Pareto_principle Where this comes into play with regards to your thought experiment... is how the larger the playing field, the fewer the large players are - they tend to get "concentrated". Plenty of Facebook alternatives exist... but why go to them? No one uses them. Large chunks of friends/family/acquaintances aren't "there" - they are on Facebook. Plenty of authors other than Steven King exist... but given a few spaces on shelves at kiosks, who's going to get those spots? You can guarantee King will get one of those few spots... The majority of activity is going to condense down until a few spots. Why? Because that is where everyone else is. That's where the skilled "creators" are. The question becomes... how do you make sure those spots are fair, open, balanced, secure, etc. Data in the hands of a few companies? What happens when they all lean left - politically? What happens when one of them gets hacked? When a "foreign actor" learns how to game that system? I've thought for the last decade that things will continue to get "worse" before they get "better". How much worse? How can we make it better? Million dollar questions.

haZard_OS 9y ago
I may be accused of nitpicking here but the Pareto Principle is, although related, distinct from the Pareto family of distributions. I don't intend to derail the conversation so, rather than typing out an explanation of obscene length, I will refer all interested parties to the following: Pareto Principle https://betterexplained.com/articles/understanding-the-pareto-principle-the-8020-rule/ https://betterexplained.com/articles/understanding-the-paret... Basics of Pareto Distributions https://actuarialmodelingtopics.wordpress.com/2017/05/21/pareto-distribution/ https://actuarialmodelingtopics.wordpress.com/2017/05/21/par... https://subversion.american.edu/aisaac/notes/pareto-distribution.pdf https://subversion.american.edu/aisaac/notes/pareto-distribu... Example of some limitations in the application of Pareto Distributions https://arxiv.org/pdf/physics/0504217.pdf https://arxiv.org/pdf/physics/0504217.pdf