Earlier quoted context omitted.
> The thing is, we have to make sure we're OK with the consequences of going this way. Would people be happy to wait for days, maybe weeks, for their video to be approved for upload to YouTube? Or they could just stop doing that and go back to how it used to be. Let people upload videos and review them only after it was flagged. Novel idea, I know.
The media companies made sure that’ll never happen again :/ That’s how we got ContentID: to prevent music and video piracy.
YouTube suspends account for linking to a PhD research on WPA2 vulnerability
101–110 of 305 posts
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#102Earlier quoted context omitted.
I think the argument is framed as wanting a human during take down dispute, not pre-approval. There's no need to treat this as an absolute. Why are human contact and reasonable expediency considered mutually exclusive? There is a lot of room in the middle. Can allow human dispute resolution for non-free accounts (i.e. paid or earning). Can limit upload frequency site wide while dispute/abuse queues are backed up (kno…
What's with the assumption here that there aren't humans involved in this decision? The linked post literally quotes Google saying: "Hi Sun Knudsen, Our team has reviewed your content, and, unfortunately, we think it violates our harmful and dangerous policy." Google has always had humans in the loop. Not for every decision but certainly the more impactful ones that affect user generated content. When I worked there,…
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#103Earlier quoted context omitted.
The thing is, we have to make sure we're OK with the consequences of going this way. Would people be happy to wait for days, maybe weeks, for their video to be approved for upload to YouTube? Would that work with the current society's attention span? Is YouTube as we know it actually feasible without AI-based moderation? Can it be a sustainable business with an army of human moderators and/or a drastically reduced sc…
I’d be shocked if the best and the brightest that Alphabet can buy can’t figure out a reputation system, manually-reviewed with AI assistance, or some other way to approve content in a convenient manner. Don’t forget also that Alphabet is swimming in money. They can certainly afford the manpower to make it relatively seamless for the user. And I really doubt users can’t be trained to associate a few hours’ wait of a…
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#104Earlier quoted context omitted.
The thing is, we have to make sure we're OK with the consequences of going this way. Would people be happy to wait for days, maybe weeks, for their video to be approved for upload to YouTube? Would that work with the current society's attention span? Is YouTube as we know it actually feasible without AI-based moderation? Can it be a sustainable business with an army of human moderators and/or a drastically reduced sc…
I’d be shocked if the best and the brightest that Alphabet can buy can’t figure out a reputation system, manually-reviewed with AI assistance, or some other way to approve content in a convenient manner. Don’t forget also that Alphabet is swimming in money. They can certainly afford the manpower to make it relatively seamless for the user. And I really doubt users can’t be trained to associate a few hours’ wait of a…
Things like timestamp of when in the video was reported might help for scanning "content" e.g. "50 people reported this 20 minute video in the 3-5 minute mark" or similar average.
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#105Earlier quoted context omitted.
> >Using algorithms to flag content is totally fine… problem is when humans cannot interact with humans anymore and AI gets to chose what is right and wrong. > Can't agree more. Simply replaced a prior system in which humans cannot interact with humans anymore and certain humans get to chose what is right and wrong. The key problem is the inability to provide human feedback more than the mechanism for making the init…
Federation is the answer at scale. You need dictatorship for effective moderation. But you also want those dictatorships small, uniform, and interchangeable so that the "citizens" of those dictatorships can easily pick the one that suits them best. Those dictatorships link up (federate) with like-minded dictatorships to benefit from network effects of sharing their content among themselves. If your content doesn't su…
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#106Earlier quoted context omitted.
The thing is, we have to make sure we're OK with the consequences of going this way. Would people be happy to wait for days, maybe weeks, for their video to be approved for upload to YouTube? Would that work with the current society's attention span? Is YouTube as we know it actually feasible without AI-based moderation? Can it be a sustainable business with an army of human moderators and/or a drastically reduced sc…
What if... instead of YouTube you'd have several intereset based sites which could actually handle their moderation and workload instead of being a giant uncaring behemoth?
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#107Earlier quoted context omitted.
What if... instead of YouTube you'd have several intereset based sites which could actually handle their moderation and workload instead of being a giant uncaring behemoth?
Who pays for the moderators? You still need a lot of manpower if you don't want algorithms, regardless of whether you have one Youtube or a thousand SpecialInterestTube.
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#108Earlier quoted context omitted.
Who pays for the moderators? You still need a lot of manpower if you don't want algorithms, regardless of whether you have one Youtube or a thousand SpecialInterestTube.
Hackaday.com does just fine. YouTube was fine before Google. Stop centralising.
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#109>Anyhow, I truly believe humanity has to rollback to operating at a human scale. >Using algorithms to flag content is totally fine… problem is when humans cannot interact with humans anymore and AI gets to chose what is right and wrong. Can't agree more.
Re: YouTube suspends account for linking to a PhD research on WPA2 vulnerability
#110Let us suppose we have a program which checks for orthodoxy according to the YouTube Terms of Service against each video submitted.
Nothing but a lack of having done the work prevents YouTube from adding to its "rejected" flag the following:
1) A list of timestarts and durations of the portions of the video violating the ToS. 2) For each item in that list, make it a tuple and add a reason -- which clause was violated?
Then outputting that to the owner of the rejected video.
Nothing stops them from doing this. Some set of words or images happened somewhere in the timestream -- the orthodoxy program has that timestamp as it churns through the video. And a specific clause was violated -- a particular word in a wordlist as an example, because there's no simple, non-composite function that just says "good or not good."
This is really not a particularly high bar as an ask.