4 ms·

Yet you often hear in this forum about how content-id is bad and is too easily (ab)used to take down content. The reality is without content-id most rights hold

by lmz 5y ago

Yet you often hear in this forum about how content-id is bad and is too easily (ab)used to take down content. The reality is without content-id most rights holders would be as screwed as Vasallo.

simonh 5y ago
I don’t think Content-ID is bad, it’s just a tool and I can see how useful it is. My problem is with the process it is a part of. Google seems to treat Content-ID as basically the whole answer, when the process clearly needs more informed discretionary moderation. The recent upheld infringement action against someone playing Moonlight Sonata is a classic example of a failure. I’ve got no problem with Content-ID matching the piano piece with the scammy upload that claimed copyright. I do have a problem with Google sticking to to guns in such an egregiously obvious case of a spurious claim of copyright. Fortunately that specific issue has now been resolved.
- tremon 5y ago
  Google seems to treat [their automated processes] as basically the whole answer, when the process clearly needs more informed [human effort] Isn't this basically their entire company policy?
  - chongli 5y ago
    You’re exactly right. Google resists hiring human moderators at every turn. I think they (reasonably) believe that it’s a slippery slope. As soon as they agree to moderate in some narrow part of their business then people will use that as a precedent to argue they should moderate elsewhere. At the end of this road is a place with so much moderation that their business model is no longer viable. We as a society have some difficult questions to face over the coming years. Do we want to surrender control over our lives to algorithms owned entirely by big tech companies?
    - selfhoster11 5y ago
      If their business model is not viable without human moderation, then perhaps it shouldn't exist. Surrendering control to "algorithms" sufficiently deeply is indistinguishable from Kafka's The Trial.
      - chongli 5y ago
        Yeah I more-or-less made this point in my other reply to this thread. I yearn for the web of my childhood (mid 90’s) when web pages were small and highly personal. People formed communities via web rings. Search was very basic but largely unnecessary because the web was so small.
        selfhoster11 5y ago
        The Indie Web and Small Web movements (I count them as distinct) have done a good job of sparking the conversation so far, plus I guess Mastodon/ActivityPub helped to some extent. I think we'll see a mini-revival of the kind of content you're looking for.
      - anchpop 5y ago
        Why shouldn’t it? Youtube has hundreds of millions of users and most of them don’t care the slightest bit about human moderation. If you only want to use platforms that have extensive human moderation systems, they exist (HN is one, I believe dang reads every comment).
        selfhoster11 5y ago
        Human-less moderation is regularly wrecking the livelihoods of many Youtube streamers, who provide valuable (often irreplacable) educational and entertainment videos to the community. Getting banned, or getting innocent videos demonetized/removed by "the algorithm" is very common. This, if nothing else, should be a motivation to provide some level of human arbitration/appeals court. I'm sure that a good argument can be made that lack of human moderation makes Youtube an unwitting vector for videos that erode political and scientific discourse by spreading harmful nonsense, but I'm sure someone else already has written about it and I don't have the energy nor will to retread that path.
Dylan16807 5y ago
Why "yet"? Those two ideas don't conflict at all. It's not like it's a binary choice between the current version of content id or no content detection at all. It is easily abused, and biased toward claimers. The fact that contentid would go too far in solving his problem wouldn't be a reason for him to complain!
anonymoushn 5y ago
It's not reassuring to read that our only choices are two different ways for small creators to constantly be robbed
simion314 5y ago
IMO the content ID stuff is not the problem but the entire system, like before you take any drastic measure have a person check, also put as input in your stupid algorithm the channel information , a 5 years old channel with many subscribers would cause a lot of noise if your stupid algorithm causes issue so even from a lazy,selfish, stupid Google dev you would put a if(channelReputition> x) ...be more careful. After working a lot with angular1 I have a really low opinion about Google devs, maybe they have some stars in their team but this does not mean that the products are good OR we can argue the devs are great but they were required intentional to build crap.
polytely 5y ago
Well content ID is bad for some cases like classical music where if you perform a piece that is in the public domain it will get claimed because the system apparently finds it really hard to distinguish different performances of the same piece.

by lmz 5y ago

Yet you often hear in this forum about how content-id is bad and is too easily (ab)used to take down content. The reality is without content-id most rights holders would be as screwed as Vasallo.

simonh 5y ago

I don’t think Content-ID is bad, it’s just a tool and I can see how useful it is. My problem is with the process it is a part of. Google seems to treat Content-ID as basically the whole answer, when the process clearly needs more informed discretionary moderation. The recent upheld infringement action against someone playing Moonlight Sonata is a classic example of a failure. I’ve got no problem with Content-ID matching the piano piece with the scammy upload that claimed copyright. I do have a problem with Google sticking to to guns in such an egregiously obvious case of a spurious claim of copyright. Fortunately that specific issue has now been resolved.

tremon 5y ago
Google seems to treat [their automated processes] as basically the whole answer, when the process clearly needs more informed [human effort] Isn't this basically their entire company policy?
- chongli 5y ago
  You’re exactly right. Google resists hiring human moderators at every turn. I think they (reasonably) believe that it’s a slippery slope. As soon as they agree to moderate in some narrow part of their business then people will use that as a precedent to argue they should moderate elsewhere. At the end of this road is a place with so much moderation that their business model is no longer viable. We as a society have some difficult questions to face over the coming years. Do we want to surrender control over our lives to algorithms owned entirely by big tech companies?
  - selfhoster11 5y ago
    If their business model is not viable without human moderation, then perhaps it shouldn't exist. Surrendering control to "algorithms" sufficiently deeply is indistinguishable from Kafka's The Trial.
    - chongli 5y ago
      Yeah I more-or-less made this point in my other reply to this thread. I yearn for the web of my childhood (mid 90’s) when web pages were small and highly personal. People formed communities via web rings. Search was very basic but largely unnecessary because the web was so small.
      - selfhoster11 5y ago
        The Indie Web and Small Web movements (I count them as distinct) have done a good job of sparking the conversation so far, plus I guess Mastodon/ActivityPub helped to some extent. I think we'll see a mini-revival of the kind of content you're looking for.
    - anchpop 5y ago
      Why shouldn’t it? Youtube has hundreds of millions of users and most of them don’t care the slightest bit about human moderation. If you only want to use platforms that have extensive human moderation systems, they exist (HN is one, I believe dang reads every comment).
      - selfhoster11 5y ago
        Human-less moderation is regularly wrecking the livelihoods of many Youtube streamers, who provide valuable (often irreplacable) educational and entertainment videos to the community. Getting banned, or getting innocent videos demonetized/removed by "the algorithm" is very common. This, if nothing else, should be a motivation to provide some level of human arbitration/appeals court. I'm sure that a good argument can be made that lack of human moderation makes Youtube an unwitting vector for videos that erode political and scientific discourse by spreading harmful nonsense, but I'm sure someone else already has written about it and I don't have the energy nor will to retread that path.

Dylan16807 5y ago

Why "yet"? Those two ideas don't conflict at all. It's not like it's a binary choice between the current version of content id or no content detection at all. It is easily abused, and biased toward claimers. The fact that contentid would go too far in solving his problem wouldn't be a reason for him to complain!

anonymoushn 5y ago

It's not reassuring to read that our only choices are two different ways for small creators to constantly be robbed

simion314 5y ago

IMO the content ID stuff is not the problem but the entire system, like before you take any drastic measure have a person check, also put as input in your stupid algorithm the channel information , a 5 years old channel with many subscribers would cause a lot of noise if your stupid algorithm causes issue so even from a lazy,selfish, stupid Google dev you would put a if(channelReputition> x) ...be more careful. After working a lot with angular1 I have a really low opinion about Google devs, maybe they have some stars in their team but this does not mean that the products are good OR we can argue the devs are great but they were required intentional to build crap.

polytely 5y ago

Well content ID is bad for some cases like classical music where if you perform a piece that is in the public domain it will get claimed because the system apparently finds it really hard to distinguish different performances of the same piece.