YouTube has announced that 72 hours of video is now being uploaded to its service every minute. Earlier this year, the statistic was that 60 hours of video was uploaded to its service every minute:

In 2007 we started at six hours [of uploads per minute], then in 2010 we were at 24 hours, then 35, then 48, and now...60 hours of video every minute, an increase of more than 25 percent in the last eight months.

This year, a 25% increase will probably take around around six months. In other words, the rate at which uploads occur is accelerating. Presumably at some point things will level off, but there's no sign of that yet, and it's not hard to see YouTube video uploads hitting 120 hours a minute or more.

Now consider the calls from some governments that Google and others pre-screen user-generated material. Just how do they think anyone can do that when every second there's one or more hours of new material flooding in? The challenge is particularly acute for video, which does not lend itself to automatic screening, unlike text, say. Such machine-based approaches are still extremely rough, and will either let through material governments want censored, or else err massively in the other direction, blocking all kinds of harmless footage.

As Google's latest figures for YouTube demonstrate, the mismatch between what governments want and what is possible is only going to get worse, thanks to Moore's Law and its analogs for storage and bandwidth. It's not clear how this is going to be resolved, but with more and more politicians calling for "something to be done", the chances of a good outcome based on rational policy making don't look good.

Follow me @glynmoody on Twitter or identi.ca, and on Google+

It's amazing just how desperate the TV networks appear to be getting. We've already discussed how they've sued Aereo, the company that sets up individual antennas to get over the air broadcasts, and then streams them over the internet to people's homes. But, now they appear to be going after Barry Diller personally for investing in Aereo. They've subpoena'd info from him directly including:

All documents concerning Aereo or the Aereo service, including without limitation (i) communications with or about Aereo, the Aereo service or Aereo’s business plans, (ii) presentations or other documents received from Aereo, and (iii) documents or communications concerning the potential impact of Aereo or the Aereo service on broadcast television networks and stations, or the producers of broadcast television.

Diller is fighting this (as you can see below) but it's a pure intimidation technique to go after those "in the fold" who invest in disruptive innovation. We saw it a decade or so ago when some major labels sued Bertelsmann -- one of their own who broke ranks to invest in the original Napster. Investors are supposed to be protected from the actions of the companies they invest in, but the dying entertainment industry folks play nasty, and they especially seem to hate it when one of their own becomes forward looking, rather than playing their myopic game. In this case, they haven't sued yet, but it's clear they're going on a fishing expedition to embarrass and annoy Diller, who (of course) came out of Hollywood in the past.

Update: Looks like Diller's already lost this part of the legal fight, which is unfortunate. Intimidating investors from funding innovative companies -- the kind of companies that actually can help move the industry forward -- isn't going to help anyone. Amazing how short-sighted Hollywood can be.

Netflix has made its share of mistakes lately in dealing with its customers, but the company's biggest problem may be that Hollywood is so jealous of its success that it's now seeking to kill off the company's service by refusing to license movies and TV shows to Netflix. This isn't a surprise. We've seen similar stories over the past year or so, but the fact is that Hollywood is so short-sighted that it's trying to hold back the tide, and in the process, causing itself more harm. Netflix isn't "the enemy." It's found a way to offer a good service that many people want and use at a price point that makes sense.

Of course, it's that part that makes Hollywood freak out. They fear the "threat" of new business models that undercut their legacy deals, and that means they want Netflix to boost prices, put even more annoying limitations on use and greatly delay and limit selection—because they stupidly think this will drive more people to the more expensive offerings from the studios themselves. The're wrong. All this does is drive more people to piracy, while killing off one of the few services that was allowed (if briefly) to effectively compete with "piracy" by offering a better overall service.

The problem for Hollywood doesn't appear to be piracy. It looks like it's Hollywood's own fear of piracy that is leading it to make really short-sighted decisions.

We recently wrote about how different parts of the music industry -- the RIAA, NMPA and DMA -- had come to an agreement on new royalty rates, as well as designating royalties for "new classifications" of services. While the groups celebrated this solution for being "flexible" for new providers, the details suggest a different story. We already expressed concerns about what are apparently licensing requirements for services that shouldn't need any license (i.e., personal music lockers). However, that was just based on the press release. When you look at the full details (pdf and embedded below), it gets even more troubling -- to the point that the whole agreement should probably be rejected.

Here's the big concern. This is a settlement among a few parties, who certainly don't represent the entire industry. Yet, if the Copyright Royalty Board and their (typically out of touch) judges accept the settlement, the details of the settlement become law. And that's problematic, because this thing is pretty crazy with restrictions -- some of which are nearly impossible to understand. If you think the tax code is confusing, you haven't tried to figure out what you have to pay to license certain services. Let's just say you want to set up a locker service that allows users to buy music which automatically goes into the locker. Well, among a ton of other rules, try this sucker on for size:

In the case of a purchased content locker service, the percentage of subpart C service revenue applicable in step 1 of &sec;385.22(b)(l)(i) is 12%. For the avoidance of doubt, paragraph (l)(i) of the definition of subpart C service revenue shall not apply. The minimum for use in step 1 of &sec;385.22(b)(l)(ii) is the appropriate subminimum as described in paragraph (b) of this section for the accounting period, where the all-in percentage applicable to &sec;385.23(b)(l) is 18%), and the sound recording-only percentage applicable to &sec;385.23(b)(2) is 22%, except that for purposes of paragraph (b) of this section the applicable consideration expensed by the service for the relevant rights shall consist only of applicable consideration expensed by the service, if any, that is incremental to the applicable consideration expensed for the rights to make the relevant permanent digital downloads and ringtones.

That's on page 40 of 44 pages. And is just one paragraph. Good luck figuring out the rest of the rules without a cadre of lawyers (oh wait... perhaps that's the idea).

But the bigger issue is that this agreement is a way to actually sneak DRM into copyright law. While existing copyright law has anti-circumvention rules, it makes no statement on how DRM actually impacts royalties or requirements (beyond anti-circumvention). Yet, this "agreement" has multiple sections that define types of DRM and with different rules for those specific cases. That is, the agreement defines the idea of a "limited download."

Limited download means a digital transmission of a sound recording of a musical work to an end user, other than a stream, that results in a specifically identifiable reproduction of that sound recording that is only accessible for listening for—

(1) An amount of time not to exceed 1 month from the time of the transmission (unless the service provider, in lieu of retransmitting the same sound recording as another limited download, separately and upon specific request of the end user made through a live network connection, reauthorizes use for another time period not to exceed 1 month), or in the case of a subscription transmission, a period of time following the end of the applicable subscription no longer than a subscription renewal period or 3 months, whichever is shorter; or

(2) A specified number of times not to exceed 12 (unless the service provider, in lieu of retransmitting the same sound recording as another limited download, separately and upon specific request of the end user made through a live network connection, reauthorizes use of another series of 12 or fewer plays), or in the case of a subscription transmission, 12 times after the end of the applicable subscription.

(3) A limited download is a general digital phonorecord delivery under 17 U.S.C. 115(c)(3)(C) and (D).

Is this really what we want in the law? A specific legal definition of DRM that applies to others despite not being a part of the negotiations? The main issue is that this is a standard contract between private parties. That's fine if it only applied to those parties who were subject to the negotiation. But thanks to the CRB process, the end result may be to stuff this private contract between private parties directly into the law, and that will expand copyright in highly questionable ways.

There are lots of other concerns about the document as well. It has lots of "this or that" calculations -- all of which default to the "greater of" option -- meaning that the rates are going to keep going up. Also, the agreement repeatedly defines "minimums" but we'll be waiting a looooooong time for you to come back with where the "maximums" are. The whole thing is crazy confusing, and while it may be perfectly fine if it were just a contract among a few players, the second it becomes part of copyright law, we should be concerned.

You probably recall all the excitement that went around when a group finally won the big Netflix $1 million prize in 2009, improving Netflix's recommendation algorithm by 10%. But what you might not know, is that Netflix never implemented that solution itself. Netflix recently put up a blog post discussing some of the details of its recommendation system, which (as an aside) explains why the winning entry never was used. First, they note that they did make use of an earlier bit of code that came out of the contest:

A year into the competition, the Korbell team won the first Progress Prize with an 8.43% improvement. They reported more than 2000 hours of work in order to come up with the final combination of 107 algorithms that gave them this prize. And, they gave us the source code. We looked at the two underlying algorithms with the best performance in the ensemble: Matrix Factorization (which the community generally called SVD, Singular Value Decomposition) and Restricted Boltzmann Machines (RBM). SVD by itself provided a 0.8914 RMSE (root mean squared error), while RBM alone provided a competitive but slightly worse 0.8990 RMSE. A linear blend of these two reduced the error to 0.88. To put these algorithms to use, we had to work to overcome some limitations, for instance that they were built to handle 100 million ratings, instead of the more than 5 billion that we have, and that they were not built to adapt as members added more ratings. But once we overcame those challenges, we put the two algorithms into production, where they are still used as part of our recommendation engine.

Neat. But the winning prize? Eh... just not worth it:

We evaluated some of the new methods offline but the additional accuracy gains that we measured did not seem to justify the engineering effort needed to bring them into a production environment.

It wasn't just that the improvement was marginal, but that Netflix's business had shifted and the way customers used its product, and the kinds of recommendations the company had done, had shifted too. Suddenly, the prize winning solution just wasn't that useful -- in part because many people were streaming videos rather than renting DVDs -- and it turns out that the recommendation for streaming videos is different than for rental viewing a few days later.

One of the reasons our focus in the recommendation algorithms has changed is because Netflix as a whole has changed dramatically in the last few years. Netflix launched an instant streaming service in 2007, one year after the Netflix Prize began. Streaming has not only changed the way our members interact with the service, but also the type of data available to use in our algorithms. For DVDs our goal is to help people fill their queue with titles to receive in the mail over the coming days and weeks; selection is distant in time from viewing, people select carefully because exchanging a DVD for another takes more than a day, and we get no feedback during viewing. For streaming members are looking for something great to watch right now; they can sample a few videos before settling on one, they can consume several in one session, and we can observe viewing statistics such as whether a video was watched fully or only partially.

The viewing data obviously makes a huge difference, but I also find it interesting that there's a clear distinction in the kinds of recommendations people that work if people are going to "watch now" vs. "watch in the future." I think this is an issue that Netflix probably has faced on the DVD side for years: when people rent a movie that won't arrive for a few days, they're making a bet on what they want at some future point. And, people tend to have a more... optimistic viewpoint of their future selves. That is, they may be willing to rent, say, an "artsy" movie that won't show up for a few days, feeling that they'll be in the mood to watch it a few days (weeks?) in the future, knowing they're not in the mood immediately. But when the choice is immediate, they deal with their present selves, and that choice can be quite different. It would be great if Netflix revealed a bit more about those differences, but it is already interesting to see that the shift from delayed gratification to instant gratification clearly makes a difference in the kinds of recommendations that work for people.

If anyone has gotten the streaming sports thing right in the past, it's been NCAA Basketball. I've actually pointed to them as an example in the past of how the rest of the sporting world should embrace streaming their games and advertise the hell out of the experience. Rather than locking down access to viewership, like the MLB, NBA, NFL, and NHL, the NCAA Basektball Tournament got it right, partnering with broadcasters to stream the games to their sites free of charge, with all the advertising one would expect on television. It was a wonderful method for ensuring that one of my monitors at work had their games on. It made it certain that I'd be that weird guy staring at images on his phone while taking the train home from work. It confused my neighbors as to how I could somehow simultaneously grill steaks on my deck while announcers went ballistic over last second hail mary shots as time ran out on the clock. I was engaged (sorry Mom and Dad, I mean engaged to the game).

And now they've taken that all away from me. To be fair, they're still streaming the games. And they promise that the streams will work across multiple platforms (computer, iPad, iPhone, Android devices, etc.). And it only costs $4.99 for the entire tournament.

But that's where the lie in all this is. It's not just $4.99 to watch the tournament on every device other than my television (where it's FREE!). There are serious mental transactions to consider here. Like most red-blooded Americans, I'm in an NCAA office bracket. Something like half the games in the tournament take place during work hours across the country. Most of us don't have televisions at our desks, in our cubicles, wherever. Watching at work is kind of the whole point here, with all the ducking and dodging from our bosses we have to do as a result. Now, we can argue all we want whether watching the tourney at work is productive or a good idea, but from the NCAA's standpoint, they shouldn't care at all. They should want people to watch. Asking them to pay $4.99 to do all of this is a massive fail, particularly since all of those same advertisements that were in place remain.

So...they're charging for something that used to be free...without adding any benefit. I watched the games on their feeds the past couple of years. This year, not so much. The obvious question is does the money they're making from the $4.99/subscriber outweigh the eyeballs that are no longer watching the advertising because of that cost. Maybe those in the comments can change my mind, but I'm fairly certain there's more folks like me out there than people buying the "package".

One of our most vocal (yet anonymous) critics posted an off-topic comment on a totally unrelated story mocking us for not having covered the story of how Hollywood has been saved (saved!) thanks to the shutdown of Megaupload. Of course, the reason we hadn't covered the story was because we didn't know about it. He referenced a couple of French news reports, which I hadn't seen until I had some time just now to catch up on some old comments. He could have submitted the stories, but he insisted that it would be a total waste of time because we ignore any story that we disagree with. That's pretty funny, considering many, many of the stories here are ones that challenge our views. And, I'm especially interested in reports of actual data, even if it conflicts with other data we've seen in the past. In fact, I'm especially interested in such stories, because my focus is figuring out what's really happening and understanding what's actually best for culture and society. So data that actually challenges my assumptions is some of the most useful data around.

Unfortunately, the stories don't actually have any data. The first one is totally anecdotal, concerning a guy who now goes to the movie rental store to rent (physical) movies since the shutdown of Megaupload, and the second one is about claims from TV stations that their websites are getting more traffic since the shutdown of Megaupload. Both are anecdotal, not data based, and really just seem to scream out the obvious: man, does the entertainment industry do a horrible job giving consumers what they want. If Megaupload's website was offering a better experience than the TV networks' own offerings... they should be firing their web designers and starting again. As for the guy renting videos... we'll see how long that lasts.

The same commenter then left a comment with some bullet point claims. He doesn't source them, so I have no idea how accurate they are, but they seem to indicate the same point in that second article. In the few weeks after Megaupload got shut down, TV station websites got more traffic. Again, this seems to simply show how bad a job they must have been doing in the past to attract viewers to their own websites. It also means that as soon as something better comes along (and it will), people will split. Ignoring that the market is telling you something is no way to build a long-term business.

Of course, we've also seen significant other data from firms that have access to a much wider view of internet traffic, which suggested that Megaupload traffic pretty quickly shifted to other, similar sites. Of course, some of those sites have since shut down or changed models, but anyone who thinks that more such sites aren't springing up (with some keeping a lower profile) simply hasn't been paying much attention to how the internet works.

Update: Netflix denies this story, though the reporter stands by it. See update at the end.

We covered how Sony Music UK jacked up prices on Whitney Houston's music minutes after her death -- then changing them back and apologizing. However, in an even more extreme case, it appears that whoever holds the copyrights on the Whitney Houston movie, The Bodyguard has pulled those rights from Netflix, where it had been streamable (found via Karl Bode, but kudos to Dan McDermott who noticed the problem and found out the details from Netflix). The reasoning is that they figure lots of people will want to buy it now, and this is a chance to cash in on her death:

Netflix rep: "Okay Dan, I just went and talked to my main supervisor as to why the movie had been pulled and the reason it was pulled was the production company pulled the streaming rights from us because all the publicity after Whitney Houston's passing there was an opportunity to make really a very large amount of money on the DVD sales of her movies. So they're going to pull all the streaming titles we have of Whitney Houston so they can make more money off the DVD sales of her movies."

Now, watch the copyright holder complain that there's too much infringement of the movie as well...

Update: Apparently Netflix is denying the story though McDermott -- a long time reporter stands by the story. Netflix claims that the streaming rights to the movie went away last year when a licensing deal ended (and it is true that Netflix has lost some streaming rights in the last few months, though I have no idea if this is one of them). However, McDermott insists that he got the story from Netflix directly. As he told Andrew Couts at Digital Trends:

“I publish three newspapers and first started in news when I was news director at WLVA in 1987. I was aware of the sensitive nature of the story and was cautious and responsible,” McDermott told us via email. “The quote I printed is accurate. I cannot speak to whether the Netflix representative was telling me the truth but I asked him to verify what the Netflix users were saying (that it was pulled after her death) and the guy came back and said what he said. I tripled checked to get the quote accurate.

“He said that he had checked with two supervisors and that the ‘main’ one told him why it had been pulled.

“Personally I believe that the kid told me what his supervisors said. I can’t imagine that they were pulled after her death in some bizarre coincidence.

“Also, it is important to note that Netflix is not the bad guy in this. Unless they lie now.”

I guess it's possible that Netflix is right, and there was confusion on the part of the supervisors...

Update 2: Indeed, it looks like my guess was correct. Netflix was right, and the supervisors of the customer service guy were wrong. Dan McDermott has admitted that the report the guy gave him appears to be wrong. He reported it correctly, but Netflix staffers gave him incorrect info. The movie was pulled from streaming back in January...

Where's ICE when you need them? For a while now, ICE has been making the case that sites that merely link to streaming content -- especially sports content -- are the most evil kinds of criminals around. It's why they're trying to extradite Richard O'Dwyer from the UK to have him face criminal charges. It's why they recently arrested Yonjo Quiroa, who will soon be tried on criminal charges as well. Both ran sites that linked to streams of TV content, including sports content. So clearly, this is serious, criminal business. In fact, if you believe the propaganda from the RIAA, these sites contribute to gangland violence and terrorism.

But, of course, when those major label guys want to set up an illegal public performance of an NFL game, where do they turn? You guessed it! One of those rogue sites.

Jason Kincaid over at TechCrunch has the rather insane story of how Vevo, the music video streaming company created by Universal Music and owned by Universal Music and Sony Music, had a booth set up at Sundance a few weeks back, where they, quite clearly, live-streamed an NFL playoff game. Yeah. In their "lounge," they had computers showing the game, apparently sourced from ESPN America -- which isn't available in the US. And the stream came from TuTele.tv, which appears to be quite similar to many of the sites that have been seized and shut down. In fact, it sounds pretty similar to Rojadirecta, which is currently fighting the US Justice Department in court.

And the folks from these major record labels used this source to live stream the football game throughout their lounge area at Sundance. It gets even more ridiculous when you remember that Congress has been trying to pass bills that would make such streaming a felony in and of itself. Senator Amy Klobuchar's streaming felony bill, S.978 -- the one that would potentially put Justin Bieber at risk -- as well as SOPA, both had provisions that made public performance of infringing content a potential felony. Those were mainly supposed to be directed at sites that allow streaming, but I think people would find it hard to argue that what happened in that bar was not a "public performance."

Of course, what this shows is that these issues are never as black and white as the RIAA would have you believe. And, just like many others, when the industry doesn't give them a convenient way to do what they want, even the RIAA's strongest supporters stoop to making use of rogue sites to potentially do "criminal acts."

I eagerly await ICE moving in to arrest Vevo execs for this blatant criminal activity.

With ICE seizing a bunch of websites that were planning to stream this year's Super Bowl, plenty of people had their ears perk up when Patriots starting quarterback Tom Brady admitted that he watched last year's Super Bowl via an unauthorized streaming site online. While his view this year was a bit more up close and personal, lots of people were talking about how NBC Universal finally decided to offer up an official stream of the game -- about a decade later than it should have done so. Of course, reports from users were that -- in typical NBC Universal fashion -- the experience was dreadful. Even more ridiculous? The livestream didn't show most of the commercials because apparently NBC Universal wanted them to pay extra to do so. So, instead, those who watched online "received a heavy dose of the same ads over and over again." In other words, the online experience was worse. Of course, since they have less competition (thank you US government!), NBC has little incentive to improve the product, so expect it to suck for a few more years now that they've finally decided to show the game online.