Monday, April 9, 2018

What do Amazon Music family subscriptions and cat trees have in common?

Hi Jeff,

Check out the attached message I got from Amazon over the weekend! My mom purchased a cat tree using my "shared payment method" for my Amazon Music Unlimited family plan.

What does a cat tree have to with Amazon music? Absolutely nothing. But somehow, starting this weekend, Amazon has decided that anyone on my family music subscription can start ordering stuff on my credit card. As explained to me by one of your helpful associates, my mom, my sister, my kids can start loading up their carts with cat trees, Fire tablets, power tools, SpongeBob T-shirts, and even that MacBook that my teenaged daughter has always wanted, using this shared payment method. Great!

But actually, it's not so great. When I signed up my family for Amazon Music Family subscription more than a year ago, the intention was to allow my mom to listen to her favorite operas, my dad to listen to jazz, my kids to listen to Drake, and my sister to listen to Hootie and the Blowfish. There was no intention to let them order whatever they wanted on my credit card, and indeed, up until this past weekend there were no purchases billed to my account. The only person that can "share" my payment method to order stuff is my lovely wife, who I explicitly granted permission to make purchases on my shared Amazon Prime Visa through "Amazon Households."

Let me tell you, it was interesting contacting your support people about this issue. One agent said one of my mother's purchases could be refunded, but "you'll need to send us an email from your mother's account only as per the company's policy." In other words, I have to go to her house to compose an email, or try to explain to her over the phone how to do this? Let me tell you, she'll have no idea what I'm talking about.

When I asked another support person why this was happening all of a sudden after more than one year in the subscription plan, he claimed it was because it was because she had ordered a digital item. When I pointed out that cat trees aren't very digital, he said he would refund that, but going forward to prevent unauthorized purchases the only option would be to remove her from the account. In other words, the support burden is now being dumped on me to not only clean up the mess but explain to my mom that the Amazon Echo I got her for Christmas a few years ago will no longer be able to play Pavarotti or Miles Davis on command.

Don't get me wrong. I love my mom, and want her to be able to buy things for kitty. But she already has an Amazon account, and her own credit card. She doesn't need to have access to my Amazon Prime visa to buy a cat tree or waffle maker or whatever.

In conclusion: I've been an Amazon customer for 20 years and have bought tens of thousands of dollars worth of stuff from Amazon. I let your devices into my home, subscribe to Prime and Amazon Video, and use Amazon Business for my own small business. I like these services and want to continue using them. But when it comes to sharing important personal data or payment methods, control needs to reside with me, the account owner. As far as I'm concerned, the only person who should have the right to use my payment information is my wife, who I explicitly granted permission via Amazon Household.

So please: Figure out a way to make Amazon Music family subscription to work as a simple way for me to share my love of music with my family members ... and let me my mom order cat trees on her own dime.

Sincerely,

Ian

Amazon Music Unlimited family plan subscription
The advertisement for Amazon Music Unlimited. Nowhere does it mention sharing credit card information.

Tuesday, March 27, 2018

A late review of Blade Runner 2049

Friends have asked me about Blade Runner 2049, the sequel to the ground-breaking 1982 film by Director Ridley Scott and a very talented cast, crew, and composer. I was 13 when the first film came out and it made a huge impact -- it was a dystopian future, to be sure, but one that was fascinating and different than the "clean" utopias depicted in Star Trek or the alien-infused space opera of Star Wars.

Back to Blade Runner 2049. I saw it on IMAX and went on an afternoon when I knew there wouldn't be many other people in the theater to disturb my experience. The film was beautifully shot, and had some incredible scenes, including the climactic, desperate fight scene outside the city's sea wall. I really liked the score by Hans Zimmer, too. He had a tough job, not only being brought in to replace the late Jóhann Jóhannsson, but also living up to the incredible soundtrack from the first film by Vangelis (which has perhaps the most dedicated cult following of any film soundtrack, judging by the mods and remixes created by fans, such as the Esper Edition. I've included one of them below).

Nevertheless, there were some serious flaws in BR49. Ryan Gosling had the lead role, and was an obvious pick (Hollywood always goes for leading men with strong "brands," which it had done when Harrison Ford in the first film). But I believe Gosling was the wrong person for the role -- he was too emotionless and almost mechanical in the fight scenes. Yes, he's a synthetic human, but considering they were designed to be "More Human than Human," I expected more humanity in Gosling's character in the 2049 generation.

There was another element that bothered me. Los Angeles 2049 has clear Asian influences, including Chinese and Japanese signage everywhere, and a cafeteria that has Asian dishes and themes. Yet there were hardly any Asian people in any roles, even extras.

Taking things one step further: Assuming demographic trends persist, in the far future society "minorities" will be white and the population will be mostly black, brown, yellow, and multiracial. Yet every major and most minor roles in Blade Runner 2049 are white, except for Olmos (reprised from the first film for about a minute), the east African shopkeeper (another short scene), and the junkyard orphanage owner.

The 1982 film was far better in this respect, imagining a future where society did include lots of people from different backgrounds and cultural influences from across the Pacific -- and even a street language based on Japanese, German, and English! 

Blade Runner 2049 isn't the only science fiction film that gets the future humanity wrong. In my long experience watching science fiction movies, only Luc Besson's 1997 film The Fifth Element tried to get humanity right, at least with secondary characters and extras who were black, mixed race, and Asian.

Despite all of this, I would like to see Blade Runner 2049 once more. I had trouble following some of the plot threads and details, and would like to see it at least once more to wrap my head around them and help me fathom the meaning of the story.

Saturday, March 3, 2018

Facebook sells out to China, and what it means for critics abroad

The New York Times published an article this week titled China is trying to police what people are saying about it around the world. I've written about China and its paranoid information-control efforts for years, but the activities to tamp down voices of dissent abroad are extreme. What I'd like to point out here are the reasons why Facebook and others are actually following China's Orwellian wishes. Take this example from the article: 
Facebook suspended Mr. Guo’s account. In a statement, the company said the account published the personal information of others without their consent, which violated the platform’s policies.
This is not just a case of China citing "hurt feelings" or some supposed rule-breaking on Facebook.
Zuckerberg is desperate to be in China. He's learned Mandarin, plays up his family connections (via his wife's family), and puts on the charm when visiting China. He's created relationships with officials at every level. His representatives have talked with companies which may be partners or acquired at some point.

Mark Zuckerbook in China
Mark Zuckerberg speaking in Mandarin during a 2015 visit
So it does not surprise me one bit that Facebook will bend over backwards to accommodate officials on censorship demands. This is the cost of doing business in China for a tech company. Media companies learned this decades ago -- read up on how Rupert Murdoch tried to operate in China. It's quite fascinating, worthy of a soap opera-style biopic.

Now it's Silicon Valley's turn, especially companies that specialize in communication -- social networks, mobile phone manufacturers, companies that sell networking hardware or cloud services, etc. The Chinese government's survival depends on information control, and if they can't do it themselves, then they force local and foreign companies to do it for them.

Don't think that this issue will be limited to human rights activists. It doesn't matter who you are -- your social media comments, family relationships, WeChat and Facebook messages, work connections on LinkedIn, and other digital footprints that reflect your attitude or influence will be used to build a social media score that determines your level of access to the country and ability to do business in China.

Tuesday, February 13, 2018

Facebook's failed march to video

Facebook has a lot of problems on its plate right now, but one of the worst trends has been the poorly thought out march to video. It came from Zuckerberg and was repeated by various senior Facebook  executives, and then became a mantra for publishers. Many publishers made large investments in video programming for Facebook, only to find the money isn't there and Facebook has decided to demote publisher content.

This is not just large media companies that got burned, I know some smaller producers who believed the "five years the feed will be all video" baloney and shifted their efforts accordingly. Some were doing really important work, too, around causes or local news. What a waste.

Meanwhile, overall video consumption in Facebook is declining, and it's the one percent that get the most engagement, regardless of who the publisher is.

I personally would not mind if Facebook turned back the clock 10 years, when most conversations seemed to be personal and text-based, and the truckloads of memes, ads, gaming achievements, real and fake news, and videos of cats playing the piano had yet to be shoved down Facebook's maw. Old-school text discussions seem to work for Hacker News and large swathes of Reddit; why can't it work for Facebook?


My Facebook feed: Ads, viral videos, and other crap I would rather not see:



Facebook feed video and ads

Monday, November 20, 2017

When lawyers are in charge of removing scraped content

Something I wrote was illegally republished on a website. This happens quite frequently with the books I publish, but usually on low-grade torrents or foreign scrapers. I can't do much about those, except to notify Google with a DMCA request to delist the copied material in search results.

However, when the infringing website is a gigantic for-profit educational company, it's a different story. I don't bother with Google, I go straight to the people running the site and demand they take it down.

The content in question was an MIT academic case published as creative commons noncommercial (specifically, Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 Unported License). The educational company has a legal email address for these types of things, so I sent a standard DMCA notice, enclosing the link of the original company as well as the infringing link.

What followed was a bureaucratic mess. The corporate lawyer insisted that I fill out some form and sign it. I refused, explaining that I didn't have to enter into a contract with the company, all I had to do was file a DMCA notice, which I had already done. This is the law of the land, and they have to follow it.

The corporate lawyer came back with this:

We are happy to remove the link to your work from our site after receiving a valid DMCA notice.  Unfortunately, your notice below is not valid under Section 512 of the DMCA. 
The DMCA Section 512(c)(3)(A) requires that when reporting copyright infringement, the notice for it must contain the following elements:
  • A physical or electronic signature of a person authorized to act on behalf of the owner of an exclusive right that is allegedly infringed.
  • Identification of the copyrighted work claimed to have been infringed, or, if multiple copyrighted works at a single online site are covered by a single notification, a representative list of such works at that site.
  • Identification of the material that is claimed to be infringing or to be the subject of infringing activity and that is to be removed or access to which is to be disabled, and information reasonably sufficient to permit the service provider to locate the material.
  • Information reasonably sufficient to permit the service provider to contact the complaining party, such as an address, telephone number, and, if available, an electronic mail address at which the complaining party may be contacted.
  • A statement that the complaining party has a good faith belief that use of the material in the manner complained of is not authorized by the copyright owner, its agent, or the law.
  • A statement that the information in the notification is accurate, and under penalty of perjury, that the complaining party is authorized to act on behalf of the owner of an exclusive right that is allegedly infringed.
[REDACTED'S] DMCA form provided to you at the link below is set up to address all six of the required elements set forth above. You may use that form, or you can provide this information in another format so long as it contains all of the required elements. Upon receiving a valid notice from you, we will promptly remove the link.
This was ridiculous. I had already made the required DMCA statement including the elements cited (link, oath, address, etc.). The only missing element was the identity of the person responsible. The scraped academic case did not say who had posted it, so as far as I was concerned, it was the company itself that decided to scrape the MIT website with the academic case against the terms of the license.

Then I thought: What if someone contacted me about suspected infringing content on a site that I run, such as the site used to sell how-to guides (sometimes mistaken for "Dummies" books)?

I wouldn't make an overpriced lawyer deal with the proboem. I'd have a production assistant take it down, or do it myself. It takes literally a few minutes to do this.

The educational company lawyer was either incompetent, trying to run up her hours, or just messing with me. I bypassed her, writing instead to the CEO of the company:

I've worked in media for a long time. If someone were to send me a complaint that illegally copied material had been posted on a site that I own or manage, and the victim contained verifiable links to the original copyrighted material and the illicitly copied version on my site, my first thing I would do would be to remove the material immediately. The second would be to issue a sincere apology to the victim. The third would be to determine what process or person was responsible, and to take steps to ensure that it never happens again.

[REDACTED] can't even get step #1 right.

Instead of having expensive attorneys in charge of this relatively straightforward and mundane process -- and still screwing it up -- how about getting some people (or more effective technologies) who can not only get things done, but can also do the right thing by victims and put policies and processes in place to ensure it's not a problem in the future?

By all means, bring in the big legal guns and the bureaucracy when it's warranted. But for basic stuff like this? You know you can do better.

Sincerely,

The illegally scraped content was removed a week later.


Saturday, October 21, 2017

Amazon cracking down on newsletter recommendation services?

A lot of self-published authors have recently noticed a problem with some of the promotions they've been using to boost sales and ranking of Kindle ebooks on Amazon. Once-safe promotional newsletters like Bookbub have apparently resulted in "rank stripping" and accusations from Amazon that the authors responsible have been using forbidden methods to boost their rank - a hallmark of the "botters" who use illicit methods to take over Amazon categories and rank in cash.

This is a big deal for legitimate authors, because they depend on these services to increase their visibility and sales. It's also a big deal for Bookbub and competing services, because they not lose sales from potential authors (a placement in the newsletter can cost hundreds of dollars) but also because they lose out on Amazon Associates affiliate income. Some of the newsletters have hundreds of thousands or millions of subscribers, who like the links to download free or discounted Kindle books.

Why is this happening? Lots of people seem to think Amazon is incompetent, or its automated systems for flagging botted books is set too high, and is throwing out the baby with the bath water:
Amazon desperately needs to be some people on this and stop relying so heavily on bots. The whole situation illustrates the dilemma of trying to beat scammers mostly by automated means. It simply can't be done. Either a lot of scammers keep on scammer with impunity, or innocent authors get caught in the anti-scam measures--or, as seems to be the case now, both things happen at the same time. 

I am going to offer an alternative hypothesis about why this is happening:

Amazon's prohibitions against "rank manipulation" have been recently expanded to include unauthorized promotional tools, regardless of whether authors consider them to be legitimate.

I don't have direct evidence of this, but I believe the following activities by Bookbub et al likely run counter to Amazon's own priorities:
  1. Maintaining a giant list of Amazon customers
  2. Encouraging behavior that runs counter to Amazon's own navigational and promotional tools
  3. Diverting money away from potential Amazon Marketing Services (AMS) spending
  4. Running up affiliate payouts
  5. Providing potential cover for actual botters and scammers, who mix in among legit authors and hope they won't get noticed.
We all know that Amazon values iron control over its platform, and wants to extend its power. Newsletters, which are basically mini platforms that depend on Amazon to survive, may be causing too many problems, and not generating enough value in return.

Does Amazon really care if Bookbub et al loses business, especially if skittish authors decide to stick with AMS? I suppose you could argue that these newsletters increase sales, but from Amazon's POV those sales might have happened anyway -- and in a more profitable way for Amazon -- if customers just used Amazon search, Amazon recommendations, Amazon sponsored ads, and Amazon newsletters to find good deals.

That's my hypothesis, anyway. And I have to admit there are a few things that go against it:
  1. If Amazon really wanted to crack down on newsletters, it could do so in ways that don't rile up the author community and increase resentment. A change in stated policy ("you can't use affiliate links or newsletters to increase sales") or cease & desist letters to newsletter operators based on real or supposed TOS violations could effectively end these types of promotions in a very short period of time.
  2. Customers really like the newsletters. To Amazon, customers are king, so pissing them off is not a good idea.
Nevertheless, if I were Bookbub (or any one of the other services that have a large Amazon customer list) I would be concerned. A business model that depends on a larger platform to succeed is incredibly vulnerable to being shut down or seriously harmed if the platform decides to change course. We've seen it happen before when Facebook cracked down on social games in 2010-2011, as well as Twitter a few years later when it decided third-party services built on the platform had to be reined in. Amazon can and will do the same if it feels newsletter promos are causing problems.


Monday, October 9, 2017

Google Adwords bans my ad, citing "Revisionist Concepts" and "Anti Content"

Google appears to be getting ready for re-entry into the People's Republic of China or Soviet Russia, judging by the "Revisionist Concepts: Unacceptable Anti Content" notification I received from the Google Adwords team for one of my AdWords ad campaigns for Lean Media.

Here's what the advertisements looked like. I created four simple text ads promoting the book and the Lean Media framework:
Anything seem wrong? These ads are about as milquetoast as you can get. Yet AdWords reviewers or algorithms thought the ad contained extremely offensive ... something. See for yourself: 
The message says "Unacceptable Anti Content ... Google policy does not permit the advertisement of revisionist concepts."

What does that even mean? The email doesn't explain, and Google's support website was no help either. Honestly, it sounds like something a Communist Party official might hand down to a hapless journalist or dissident before sentencing him or her to the gulag.

To be clear: Lean Media is a production framework for people who create media. The 50,000-word book I wrote is apolitical, non-offensive, and encourages media creators to make better media that audiences love. There are tons of real-world examples, from an art-photography publisher in China to global media brands like Minecraft and The Simpsons. Even if the book were related to politics or some other controversial subject, there's no reason Google should ban it ... at least not in this country.

Putting on my tinfoil hat: What I suspect happened is my recent public criticisms of Google AdWords about misleading settings for locations and ad fraud got noticed, and someone flagged it for additional review. "Anti-content" was the only thing they could come up with, and they banned the ad.

Do no evil, indeed.