6 ms·

I would really like a recommendation engine that I own and control. It would know what I read and watch. It would find new stuff for me, dedupe it, listen to m

I would really like a recommendation engine that I own and control. It would know what I read and watch. It would find new stuff for me, dedupe it, listen to my feedback,and learn from its mistakes. It would work with all kinds of media, not just youtube or HN. And it would all be private. I don't mind paying for my content but I do mind being the product.

rvrabec 6y ago
+1 I'd also pay for something like this...
- h0p3 6y ago
  ヽ༼ ຈل͜ຈ༼ ̿̿Ĺ̯̿̿̿ ̿༽Ɵ͆ل͜Ɵ͆ ༽ﾉ ¸.••.¸¸.••.¸¸.••.¸¸.••.¸ └[∵┌]└[ ∵ ]┘[┐∵]┘ +1. I'll even pay a liquid democracy ideology subscription fee for this to be built. I will pay for 3 years without even a working prototype. Do it right, or don't do it at all; take the space by storm. Build it to give it away! I'm willing to accept a number of copyleft licenses. Iteratively, decentralize until it hurts, then centralize until it works. Take 'em down bois! Prove your progress in a P2P-hosted git. Show me it is owned by the workers, from leadership to infrastructure to what runs on my machine. You freely give away 10% gross of what you are given and show me what you are made of (when you pick out your gods). Assume your user has a Pinephone. I want it synchronized across my devices on Dat. I want it to be integrated with my Tiddlywiki and Fraidycat. Favor Rust in your construction. Make it live in both my terminal and browser as nearly dialectical equals. Help me choose my moderators (those who censor and shape), make it simple for me to re-use the models of individuals in how I remodel my filter-bubbles. I work hard to make my data public, and I will work with others who are willing to do the same; build with our data if you need. I want us to pick out what is salient together using fundamental human trust established over mere wires in Gradualized T42T. ( ͡°( ͡° ͜ʖ( ͡° ͜ʖ ͡°)ʖ ͡°) ͡°) ¸.••.¸¸.••.¸¸.••.¸¸.••.¸ ヘ( ^o^)ノ＼(^_^ )
  - jamiek88 6y ago
    Wow GPT-3 is getting good.
    - h0p3 6y ago
      I agree that GPT-3 is impressive (https://philosopher.life/#Aispondence https://philosopher.life/#Aispondence). You might agree: "thought and discussion termination isn’t cool," right? I hope you aim to live up to your own principles. Maybe you meant to say something else to the human being on the other end of the line.
  - aasasd 6y ago
    I see it's getting harder to satisfy you, by the minute.
    - h0p3 6y ago
      Oh, some people say I'm an insatiable monster. I appreciate that you have taken the time to reply. If you are really here for a discussion, I'll have one. I do care about what you say and don't say (and I'll prove it to you). I do ask quite a bit of those in power (https://philosopher.life/#Power https://philosopher.life/#Power).
hansvm 6y ago
> And it would all be private. I don't mind paying for my content but I do mind being the product. Some portions of a personalized, cross-service recommendation engine are a lot more viable if centralized (e.g. any scraping/indexing, and plausibly the personalization itself). Supposing such a service had no intention of you ever personally being the product, what would they have to do in order for you to still be comfortable handing that data over to them?
- nitrogen 6y ago
  Host in my closet or basement with a one-way data flow. In only, not out.
  - Alphaeus 6y ago
    How much of the general public is going to be capable to set up and maintain a server in their basement? To challenge YouTube, targetting just the programming crowd isn't going to be sufficient.
    - nitrogen 6y ago
      The general public seems to be capable of setting up and maintaining a Nest thermostat. The only difference would be that data only goes down, not up.
      - Alphaeus 6y ago
        I guess if you sell a pre-constructed device, that could work. There would be other issues (you have to spend money on hardware, potentially have to get it shipped, may not be able to get access to the device at all based on your region, keeping it always on and accessible, etc).
ashtonkem 6y ago
As a computer program, this is probably not feasible. While in theory you are possibly capable of writing the algorithms for your own recommendation engine using whatever ML tools are popular today, the issue you'd have is data. YouTube's recommendation algorithm effectively only works because it's capable of aggregating millions of users behavior to feed into their own system. Getting that data personally would be very difficult both because YouTube doesn't want to enable competitors, and because consumers would be extra freaked out to find out about some random engineer outside of YT getting their watch history. As a non-program, you can always depend on the means that humans have used for content discovery since time immemorial: word of mouth.
- imglorp 6y ago
  Well just for me I don't care what's trending so don't need the feed of other people's views. But absolutely needing a feed of new posts might be problem.
  - creato 6y ago
    That's really missing the point. Recommendation engines are a lot more dumb than you are thinking. They don't understand the content at all. They only know that thousands of users that watched the same videos as you in the past also watched this other new video, so you might like it as well. A recommendation engine that works solely based on your own watch history would require a totally different approach using algorithms that are years/decades/forever away from existing.
    - imglorp 6y ago
      Good points. It does make sense that the job is simpler with everyone's history and inputs, if you can keep PII out of it. Are there any good references on recommendation aggregation algorithms? Here's a simple proposal that keeps everyone's stuff anonymous and allows a rec service to pay bills. Distribute the pool of everyone's watch histories/prefs/etc using IPFS, which is a DHT with static names and some storage. User clients put the share data into IPFS and register into a public pool. The client remembers their own hash. Rec services (there can be multiple) pull everyone's sets periodically and aggregate/indexe them. Users can then retrieve a batch of new recs by sending a small fee to the service of their choice along with their hash and the server will place some new recs into their bucket. If the recs are not suitable, they can try another service.
fourtens 6y ago
What do you think of https://findka.com/ https://findka.com/ No affiliation, but it seems to be trying to do what you want
dghughes 6y ago
I know what you mean, if this is what you mean. YouTube has become so laser-focused that it's become ridiculous. I watch videos and get suggested similar videos. Eventually I'm only watching one type of video. >I don't mind paying for my content but I do mind being the product. This is another thing that bothers me. Being guilt-tripped for not watching an ad, or being a Patreon subscriber, or YouTube Red subscriber. Even if I don't watch ads or pay a fee I am, me, this individual is valuable just for who I am. My data, my habits are valuable to Google just as much as it is for me to pay a fee or to watch ads.
- SpelingBeeChamp 6y ago
  Google is not the person who created the content you are consuming. That person or group of people gets nothing from viewers who "don't watch ads or pay a fee."

rvrabec 6y ago

+1 I'd also pay for something like this...

h0p3 6y ago
ヽ༼ ຈل͜ຈ༼ ̿̿Ĺ̯̿̿̿ ̿༽Ɵ͆ل͜Ɵ͆ ༽ﾉ ¸.••.¸¸.••.¸¸.••.¸¸.••.¸ └[∵┌]└[ ∵ ]┘[┐∵]┘ +1. I'll even pay a liquid democracy ideology subscription fee for this to be built. I will pay for 3 years without even a working prototype. Do it right, or don't do it at all; take the space by storm. Build it to give it away! I'm willing to accept a number of copyleft licenses. Iteratively, decentralize until it hurts, then centralize until it works. Take 'em down bois! Prove your progress in a P2P-hosted git. Show me it is owned by the workers, from leadership to infrastructure to what runs on my machine. You freely give away 10% gross of what you are given and show me what you are made of (when you pick out your gods). Assume your user has a Pinephone. I want it synchronized across my devices on Dat. I want it to be integrated with my Tiddlywiki and Fraidycat. Favor Rust in your construction. Make it live in both my terminal and browser as nearly dialectical equals. Help me choose my moderators (those who censor and shape), make it simple for me to re-use the models of individuals in how I remodel my filter-bubbles. I work hard to make my data public, and I will work with others who are willing to do the same; build with our data if you need. I want us to pick out what is salient together using fundamental human trust established over mere wires in Gradualized T42T. ( ͡°( ͡° ͜ʖ( ͡° ͜ʖ ͡°)ʖ ͡°) ͡°) ¸.••.¸¸.••.¸¸.••.¸¸.••.¸ ヘ( ^o^)ノ＼(^_^ )
- jamiek88 6y ago
  Wow GPT-3 is getting good.
  - h0p3 6y ago
    I agree that GPT-3 is impressive (https://philosopher.life/#Aispondence https://philosopher.life/#Aispondence). You might agree: "thought and discussion termination isn’t cool," right? I hope you aim to live up to your own principles. Maybe you meant to say something else to the human being on the other end of the line.
- aasasd 6y ago
  I see it's getting harder to satisfy you, by the minute.
  - h0p3 6y ago
    Oh, some people say I'm an insatiable monster. I appreciate that you have taken the time to reply. If you are really here for a discussion, I'll have one. I do care about what you say and don't say (and I'll prove it to you). I do ask quite a bit of those in power (https://philosopher.life/#Power https://philosopher.life/#Power).

hansvm 6y ago

> And it would all be private. I don't mind paying for my content but I do mind being the product. Some portions of a personalized, cross-service recommendation engine are a lot more viable if centralized (e.g. any scraping/indexing, and plausibly the personalization itself). Supposing such a service had no intention of you ever personally being the product, what would they have to do in order for you to still be comfortable handing that data over to them?

nitrogen 6y ago
Host in my closet or basement with a one-way data flow. In only, not out.
- Alphaeus 6y ago
  How much of the general public is going to be capable to set up and maintain a server in their basement? To challenge YouTube, targetting just the programming crowd isn't going to be sufficient.
  - nitrogen 6y ago
    The general public seems to be capable of setting up and maintaining a Nest thermostat. The only difference would be that data only goes down, not up.
    - Alphaeus 6y ago
      I guess if you sell a pre-constructed device, that could work. There would be other issues (you have to spend money on hardware, potentially have to get it shipped, may not be able to get access to the device at all based on your region, keeping it always on and accessible, etc).

ashtonkem 6y ago

As a computer program, this is probably not feasible. While in theory you are possibly capable of writing the algorithms for your own recommendation engine using whatever ML tools are popular today, the issue you'd have is data. YouTube's recommendation algorithm effectively only works because it's capable of aggregating millions of users behavior to feed into their own system. Getting that data personally would be very difficult both because YouTube doesn't want to enable competitors, and because consumers would be extra freaked out to find out about some random engineer outside of YT getting their watch history. As a non-program, you can always depend on the means that humans have used for content discovery since time immemorial: word of mouth.

imglorp 6y ago
Well just for me I don't care what's trending so don't need the feed of other people's views. But absolutely needing a feed of new posts might be problem.
- creato 6y ago
  That's really missing the point. Recommendation engines are a lot more dumb than you are thinking. They don't understand the content at all. They only know that thousands of users that watched the same videos as you in the past also watched this other new video, so you might like it as well. A recommendation engine that works solely based on your own watch history would require a totally different approach using algorithms that are years/decades/forever away from existing.
  - imglorp 6y ago
    Good points. It does make sense that the job is simpler with everyone's history and inputs, if you can keep PII out of it. Are there any good references on recommendation aggregation algorithms? Here's a simple proposal that keeps everyone's stuff anonymous and allows a rec service to pay bills. Distribute the pool of everyone's watch histories/prefs/etc using IPFS, which is a DHT with static names and some storage. User clients put the share data into IPFS and register into a public pool. The client remembers their own hash. Rec services (there can be multiple) pull everyone's sets periodically and aggregate/indexe them. Users can then retrieve a batch of new recs by sending a small fee to the service of their choice along with their hash and the server will place some new recs into their bucket. If the recs are not suitable, they can try another service.

fourtens 6y ago

What do you think of https://findka.com/ https://findka.com/ No affiliation, but it seems to be trying to do what you want

dghughes 6y ago

I know what you mean, if this is what you mean. YouTube has become so laser-focused that it's become ridiculous. I watch videos and get suggested similar videos. Eventually I'm only watching one type of video. >I don't mind paying for my content but I do mind being the product. This is another thing that bothers me. Being guilt-tripped for not watching an ad, or being a Patreon subscriber, or YouTube Red subscriber. Even if I don't watch ads or pay a fee I am, me, this individual is valuable just for who I am. My data, my habits are valuable to Google just as much as it is for me to pay a fee or to watch ads.

SpelingBeeChamp 6y ago
Google is not the person who created the content you are consuming. That person or group of people gets nothing from viewers who "don't watch ads or pay a fee."