> It’s modeled off a federal law passed in 2010 that caps ad volumes on cable and broadcast TV, but doesn’t apply to streaming services.
Why did that law not apply to streaming services in the first place? The internet was very much alive and kicking in 2010. Sure, streaming wasn't as prevalent as it is today, but it wouldn't have taken a lot of imagination to see the same problem would become an issue on the internet as well.
The Internet, and before it, computers, broke our legal system. There are loads of things that we decided were bad and banned, but "thing but on computer" or "thing but online" somehow were interpreted to be exempt.
For instance, there's a law banning video rental stores from sharing customer records, because it's obviously bad if private entities are allowed to collect and use potentially private information like media consumption habits. But movie streaming? Every detail about every piece of media you read or watch, when you watch, when you pause or bounce, every interaction and speck of attention catalogued and actively used to guide consumer behavior? That's fine actually, totally allowed.
How about copyright? Right of first sale dictates that you can do whatever you want with a purchased copy of some media, short of distributing copies. You can give it away, sell it, lend it out, modify it, make personal copies, whatever. But what about "media but on computer"? That all goes out the window. Oh, you don't own a copy, you just have a non-transferable limited license to view that media on a specific device for as long as the distributor doesn't change their minds. An insane legal fiction that magically nullifies hard-won rights.
The video store example is funny because iirc, it wasn’t until someone high up/very involved in government got bit by it. During Robert Bork’s failed Supreme Court confirmation, a reporter figured out he rented porn. Maybe it was something less raunchy / embarrassing than porn but either way, iirc, they got that law on the books fast after that….
The leak was inspired by Bork's opposition to privacy protections beyond those explicitly outlined in the constitution. [0]
On September 25, the City Paper published Dolan's survey of Bork's rentals in a cover story titled "The Bork Tapes". The revealed tapes proved to be modest, innocuous, and non-salacious, consisting of a garden-variety of films such as thrillers, British drama, and those by Alfred Hitchcock. [1]
The VPPA very much applies to online entities. Netflix can collect all the info it wants about you, but is very much limited in what it can share with external parties.
If anything, the law has given cover to shady walled garden business practices that would not have survived otherwise.
You read wrongly. The 2013 amendment merely allowed customers to consent to disclosure electronically via the Internet. Before then, it had to be in writing. It didn't change 18 USC § 2710's explicit application only to a "video tape service provider", and that is how the law still reads today.
I think until now, the only real thing preserving these consumer protections (and civil rights, while we're at it) was simply the practical difficulty of collecting and compiling that information. Now it's trivial and effortless to collect this information. The old laws still apply, but now the rubber meets the road in actually having them enforced.
>"thing but on computer"
From a tech layperson, all the tech "innovation" I'm seeing seems to just be old stuff but "online" and therefore not subject to the "old rules".
How about liability for publishers? New York Times publishes something damaging and false? Liable! YouTube publishes something damaging and false but since they did it with a computer they're immune!
Now that youtube and meta and tiktok choose what is put in front of you they are the publishers. But the law, passed for the early web, is stuck in the past.
Congratulations, you fell victim to the 'platform vs publisher' liability misinformation. It doesn't work like that and has never worked like that, nor should it except for the perfidious pushers of that misinformation.
A prioritization or recommendation algorithm does not count as publication. The work was already published by somebody else. Do you blame a library card catalog for listing by subject, title, or chronology?
If a librarian put a book out on the front table with a "recommended reading" sign then yeah that seems fair for them to carry some liability if that book were actually libelous. And so it should be for recommended posts on sites like Youtube, Instagram, etc. A chronological or alphabetical index is a factual catalogue of information. A recommendation is you vouching for the material. Totally different.
> Do you blame a library card catalog for listing by subject, title, or chronology?
I would if someone reordered them based on some subjective "engagement" metric.
The card catalog is not a recommendation engine. YouTube's recommendations are... literally a recommendation engine. I think platforms should be legally liable for the things they promote via subjective choices. Pity the law isn't set up that way.
It worked like that before they changed the law and it can work like that again.
> The work was already published by somebody else.
This is just wrong. It is literally the platform that does the publishing. However, section 230 says that we won't treat the platform as the publisher.
This is not some logical necessity. It's just a law that we can change.
My guess, having only looked at the text of the law but not into any of the legislative history, is that it was for technical reasons. This is based on how they worded it. The text says it applies to "a television station, cable operator, or other multi-channel video programming distributor".
This suggests they were thinking of linear television. Some searching tells me that in fact this is how it was apparently interpreted, for when it was applied to cable TV it was not applied to on-demand cable programming. It was just applied to the regular cable channels.
With linear TV everything is prepared in advance. When they sell an ad slot they know what program they will be showing at the time. There's plenty of time to match the ad volume to the program volume, which I suspect in 2010 could not be reasonably automated.
With on-demand you don't know what programs the ad will be in until the program actually starts. You could potentially be showing that ad in thousands of different programs at approximately the same time. If the level adjustments could not reasonably be completely automated this may have been impractical.
> With linear TV everything is prepared in advance. When they sell an ad slot they know what program they will be showing at the time. There's plenty of time to match the ad volume to the program volume, which I suspect in 2010 could not be reasonably automated.
Not really. There's a lot of live programming. Ad campaigns may be cancelled and replaced close to the time of airing. Local stations and cable systems preempt national ads and insert their own ads at times.
The way this was resolved was not by tuning ads to the content they interrupt, it was by setting standard audio levels for all programming and tuning the ads to fit that standard.
Yeah you calculate ReplayGain metadata for the media and let the playback client do normalization to a target level. Unless it's live streaming, pre-calculating gain levels is a non-issue. All the CD ripping/music library software in the 00s already did this because of the loudness war.
The US government typically doesn't try to preemptively regulate things (which is getting to be a problem as it now is too sclerotic to respond quickly to developments). Streaming services in 2010 were mostly paid subscriptions with no ads, I don't think the idea was on anyone's mind.
YouTube existed in 2010, and was primarily supported by ADs. Sure at the time it didn't really show what would traditionally be considered television, but it seems like the same logic would apply.
I'm reminded of how many patents that were due to expire after their 20 year lifespan got renewed simply by adding "using the internet" tacked on at the last minute.
Why did that law not apply to streaming services in the first place? The internet was very much alive and kicking in 2010. Sure, streaming wasn't as prevalent as it is today, but it wouldn't have taken a lot of imagination to see the same problem would become an issue on the internet as well.