Thursday, April 26, 2018

"The Filtering Practices of Social Media Platforms": House Judiciary Committee holds hearing today



There is a hearing at the House of Representatives today. With the House Judiciary Committee, “The Filtering Practices of Social Media Platforms” at 10 AM today in the Rayburn building, with the main link here.   I haven’t been able to find live video coverage anywhere yet. 
   
Electronic Frontier Foundation has an article by Corynne McSherry and India McKinney, “Platform censorship won’t fix the Internet”  where EFF submitted a statement in advance (embedded PDF) here. The statement mentions an instance where some LGBTQ videos were put into restricted mode (age verification by sign in) by YouTube.   I usually won’t embed videos in restricted mode.

The EFF article suggests that it was suddenly disinvited from speaking.  I don’t know why.
  
Censorship debate certainly has focused on FOSTA and Backpage recently, but also on protecting minors (the COPA case of the 2000’s). But more recently censorship has included terror, violence promotion, weapons, fake news or election manipulation (itself an obscure issue dating back to the early 2000’s – I’ve covered it here before), white supremacy, and even some new trends like Incel (the Toronto incident).  “Gratuitous speech” and “implicit content” could come up as issues, and they are much more subtle.  (See the Facebook guidelines on Monday, April 23 posting).

Tuesday, April 24, 2018

Cloudflare, and then Vox, both double-slam FOSTA as a virus likely to destroy Section 230 and Internet user generated content



Recently, Vice reported that Cloudflare – which, remember, in a fit of virtue signaling, had banned Daily Stormer in August after Charlottesville and which pretty much got that alt-right site blackballed – now has banned “switter”, an Australian “dating” site, out of fears of the downstream liability risks from FOSTA. 

Furthermore, after an initial vague statement about TOS, Cloudflare said that the Internet is a little less free than it was because of unintended consequences, and that Congress, under political pressure, was willing to ask for some sacrifice from the elitist know-it-alls to protect the “people”.  Cloudflare especially said that Congress is lazy about understanding Internet “infrastructure.”
It’s notable that Cloudflare took this action against a company that operates out of Australia based on US laws. Imagine the parallels. 

Then Vox came along, April 23, with a double-header on the threat to Section 230.  The “nightcap” was an article by Emily Stewart, “The next big battle over Internet freedom is here”.  The obvious question, which will boil for the next “two to five” years, will be downstream liability over other dangerous issues:  terror recruiting, weapons information, fake news (itself obviously potential libel), revenge porn and cyber stalking, as well as “misuse” of consumer data.  (Targeting group-centric “hate speech” might be harder because of the way the First Amendment is set up, but Europe already does that.) But the “afternoon game” had been a Vox explainer by Aja Romano is titled more stridently, “A new law intended to curb sex trafficking threatens the future of the Internet as we know it.”  Romano notes that some groups want to take this further into areas like pornography addiction and degradation of women.


One aspect of the discussion that seems to be missing:  different providers do different things and have widely varying degrees of downstream risk, at least to sex trafficking promotion.  Websites that cater to personal needs obviously would have the highest risk.  Hosting services (which websites use but which readers don’t directly log on to) would have mathematically much lower risks, as would CDN’s like Cloudflare and SiteLock. Social media would have higher risks, but already have become proactive in screening a lot of content with artificial intelligence (which Facebook and YouTube, despite all their controversy, are getting better at). The same holds for parallel problems in the copyright world (covered by a different set of laws – DMCA Safe Harbor).  

I could back up and restart a discussion of personal moral philosophy, about the “privilege of being listened to”, as I have in my books.  No doubt, user generated content (like mine) can hold big media and politicians honest, but the weakness of this argument seems to be rooted in the obvious history, that until the mid 1990s at the earliest, you generally couldn’t publish yourself “to the world” unless the publication could pay for itself, and we got along and survived Vietnam, Nixon, and got as far as starting to debate gays in the military in old fashioned media, before the Internet suddenly gave us all a new voice.

No doubt, though, asymmetric individual speech can weaken group-based conventional activism, as well as, with some speakers, exacerbate social tensions over inequality – at least through the hypocrisy which others can pick up on.

We’re starting to see why authoritarian societies, even milder ones (like Singapore) believe they have to collar individualized speech to maintain public safety. 

Monday, April 23, 2018

Maybe Mark Zuckerberg really did conquer the world the way an alien would!; FB finally publishes new objectional content guidelines and tackles meta-speech



On GMA on ABC this morning, as I had a “free” vegan sausage (party) in hotel bar, I saw Aleksandr Kogan “apologize” for his “misuse” of Facebook consumer data, as in this New York Times story from Matthew Rosenberg.  Kogan’s historical perceptions and his rationalization of them at the time were interesting to me (as they would be to Dr. Phil).  The article reviews why Facebook thought it had taken care of the problem in 2015.

Then Amanda Taub and  Max Fisher explore “Where countries are tinderboxes and Facebook is a match” here, with a flaming gif illustration.

Facebook was designed for western countries with strong institutions to buffer social stability against mobs.  But even in the West, as we saw, tribalism erupted after 2014, as we saw from BLM and then the “alt-right”.   

Brian Resnick’s article on what draws people to extremism (the alt-right) even in the US is rather unsettling.



Update: April 24

Chris Wylie, the "gay vegan" software genius (there are others  -- he reminds me of Danganronpa in Reid Ewing's world) who blew the whistle, will testify before Congress soon and offers this advice on how to fix Facebook. 

Further: April 24 

Facebook has released its objectionable content policy, here.  Look at the "hard questions" regarding meta-speech under the hate speech area.

Chris Farivar of Ars Technica remarks that Facebook is finally "telling its users what's allowed". 


Friday, April 20, 2018

Will Facebook's easing into GDPR be good enough in the U.S. for most users?


Staci Krame chastises Facebook in the Sunday (April 15) Outlook section, “Facebook could easily make privacy the default. It hasn’t.” 

She wants a standard of opt-in rather than making you opt out through several screens.
  
Emily Stewart of Vox explained on April 18 what you need to know about the new privacy settings. 
  
This is supposed to comply with Europe’s General Data Privacy Regulation (GDPR).  For example, you can tell it if you want target ads.

But others say that Facebook won't fully comply with GDPR in North America, at first at least. 

It’s significant that you can opt out of facial recognition (should that be an opt-in) although I don’t see how that stops false tagging in photos (that has happened to me only once). But the facial recognition problem seems to comport with the skittishness that some people have these days about being photographed in bars, where the etiquette expected is narrower than it was say ten years ago.
   
The other deeper question is how public people want to be.  I did want to make myself public when I wrote the first DADT book because, according to the standards of the time, the arguments I wanted to make were unique enough that they demanded my own personal narrative become public and searchable to back it up.  That isn’t generally true of most people, and I don't think what I did is a model for others to follow, generally speaking.  However,\indeed, in the past few years we’ve seen a disturbing trend of people being coerced to support other people’s causes (sometimes workplace or reputation related) on their own personal social media accounts.

Tuesday, April 17, 2018

In LiveJournal case, plaintiff tries to outflank the DMCA Safe Harbor



Here’s a DMCA copyright case to watch: Mavrix Photography v. LiveJournal, as explained by Corynne McSherry at Electronic Frontier Foundation, link.  

The problem is that the plaintiff sued anyway, bypassing waiting for a takedown under DMCA Dafe Harbor, on the “theory” that the site moderated the content but did so inadequately.

  
And unfortunately the Ninth Circuit, usually favorable to free speech, is letting it go to trial.

Monday, April 16, 2018

Supreme Court takes up expansion of Internet sales tax



On Tuesday,  April 17, 2018, the Supreme Court will hear oral arguments on litigation brought by the State of South Dakota on whether the a state can collect sales taxes from residents buying online from out of state retailers where retailers don’t have a physical presence in the state.
  
The NBC News story by Peter Williams is here.

Very much as the case with legislation that has been proposed, the measure could make compliance much more costly for small businesses.


Businesses normally register to collect sales taxes only in the states that they operate. But conceivably a system could be developed to route payments among states, and could be offered by hosting providers along with payment portals to businesses.  Automatic transfer of payments is common;  I worked on mainframe systems that could do that back in the 1990s when I was “working”.

Update: April 17

Here is the transcript of the oral arguments today.

Sunday, April 15, 2018

Brass musician explores the copyright and monetization issues on YouTube when making transcriptions



Peter Opaskar plays the tuba and like to make transcriptions of music for the tuba.  He has devised a sophisticated technique of recording separate sessions of himself and superimposing them, and idea I have recently considered for the close of my own third Sonata (to be discussed in more detail later on a Wordpress blog). 

But his article in Ars Technica is important because it shows the tricky problems in trying to monetize on YouTube.  At one time he was not allowed to run ads because his transcriptions were considered copyright infringement.  Later, YouTube changed its mind on the maskup issue, especially after some songwriters indicated they had no objections (sometimes royalties had to be shared).

  
But then the composer/transcriber/performer found that YouTube had a new rule:  you couldn’t monetize until you had a minimum number of unique visitors and visits.
     
In classical music, transcriptions are sometimes regarded as legitimate separate compositions (this is most common with variation sets, or with prelude-and-fugue on a theme, often of Bach).  Some composers have said that all composition involves some copying (rather like my statement in a high school chemistry class that all learning is memorizing – in organic chemistry, maybe it is). 

Thursday, April 12, 2018

Democrats blame Zuck, Facebook for their own past laziness



Daniel Henninger puts the most convincing spin yet on Mark Zuckerberg’s testimony.  That is to say, it wasn’t Mark’s job to catch Russian bots, it was Obama’s.  Hillary Clinton didn’t lose the election because of Facebook, or because of the Comey letter;  she lost because she ran a complacent campaign and had behaved badly herself, at least in terms of showing the technical competence to run things now. And, of course, there is the Electoral College problem. 
  
Of course, Trump isn’t competent either;  and both parties now have problems attracting the right kind of talent to run for public office.  That’s a systematic failure.  You can imagine people a lot more capable of actually doing the job of being president:  Chris Cuomo, Anderson Cooper, both journalists and both rather geeky.  David Hogg turned 18 today, and he talks like he really wants to do public service.  17 years too young.  The irony is one wonders how Mark Zuckerberg would function in office.  Asperger’s syndrome in the White House?  You can imagine some business executives a lot more suited than Trump – Mark Cuban, Barbara Corcoran (both on Shark Tank), and Tim Cook.  Suddenly the idea of LGBT candidates comes up, maybe even trans (who knows national security better than Kristin Beck?)
  
  
Seriously, we really do have a problem.  Raising money for candidates seems pimpy and unwelcome.  But that’s part of our own cultural divide, the enemies can exploit.  Maybe we have to get over that. 
  
Of course we have to take into account Cambridge's claim that the data wasn't used for the election -- and now there is speculation about a possible paid version of Facebook. 
   
There is a conceptual problem with the way we have leveraged user-generated content, offered by speakers who (like me) want to be noticed and be seen as “influencers” without going through the grime of partisanship.  We sensed some of this back around 2005 when there was a flap over the idea that bloggers could be unintentionally making de facto campaign contributions with free content (not paying its own way the way normally published material did).  Since vanity publishing had not become a practicable vehicle for self-expression until the late 90s, there is a certain gratuitousness to it – apart from its challenge to working with others in a spirit of shared partisanship (or “solidarity”).  That alone sets up the possibility of combining truly hacked information (like from Equifax) on the dark web with provocative speech to target identifiable individuals – taking what the Russians did with groups (pitting them against each other) to a new, more dangerous level.  But a lot of this had been going on before Cambridge made its heist.  This sounds like a profound problem in the way we conduct our “politics” and it relates to our personal morals, our personal stake in other people.  
  
April 13:   I had a conversation today indicating that Aleksandr Kogan was the main conduit for misuse of Cambridge data, based on one personality survey he had put on Facebook

Wednesday, April 11, 2018

Tribal societies expect men to sacrifice themselves for women and children



Suzanne Fields, who has a syndicated column in the conservative paper “The Washington Times”, has a rather old-fashioned column about cis masculinity, “A Good Man Is Still Hard to Find”.   

There is an interesting quote from Sebastian Junger, which may be a little out of context (although Junger’s “The Tribe” says some of these things).

Suzanne lays it on the line when she quotes Joyce Benson’s definition of a man, as someone you can depend on in dealing with an enemy.  Men by definition have to allow themselves to become fungible within the tribe.  She means that four of the men who died in the Aurora shooting died shielding women.

To say the least, some of us have trouble dealing with this being expected when we grow up.  But that’s how it was in the days we had a male-only draft.
  
George Gilder had indulged in this kind of thinking in his book “Men and Marriage” (1986), which had rewritten “Sexual Suicide” (1973).

Now we see the idea coming back in debates on what we now expect of teachers.  But laying your self on the line has crossed tribal and familial lines in modern society.  Look at the openness to offer organ donations online, something not possible when I was growing up.
  
 Another good piece to look at is "The Anthropology of Manhood" by Sebastian Junger (March 2018) in National Review. 

Saturday, April 07, 2018

Internet "surveillance capitalism" business models in real trouble; DHS wants to monitor "media influencers" (like me?)



Back in January, HuffPost had already run a big story on why the “surveillance capitalism” business model of both Google and Facebook is in trouble, by Paul Blumenthal.    There are new rules called the General Data Protection Regulation, or GDPR, set to go into effect in May.  And we all know the “national security” threat of Russian meddling – and the controversies of the Internet and campaign finance that date back to the early 2000’s and were forgotten.

Pondering what I wrote yesterday – I can counter-ponder:  Hillary Clinton lost the election because she ran the campaign poorly, and because of the Electoral College, not just because of the “Russians”.   And users do know that the content they are fed is based on their past activity, especially likes and commenting activity (“Likenomics”). True, it drove them into bubbles.  But users should know that if they kept on liking birtherism, that’s what they’ll see in their news feeds.
All of this matters as Congress is surely going to consider restricting the use of CDA230 (Section 230) to protect social networking sites that manipulate what users see.

On the other hand, hosting providers don’t do that.  They just serve up what a user goes to (the user has paid for the domain name and hosting service, or could have provided his own).  The same with blogging platforms.

Robert Kuttner weighs in further with a more recent article on How to Regulate Facebook.
   
There is also a story, by Michelle Fabio of Forbes, that the Department of Homeland Security wants to maintain a database if journalists and “media influencers”.  Maybe that includes me!!   Count me in.  I was willing to allow my likenomics to be sold in order to have influence (and escape having to be loyal to other groups to be heard).   Beyond what has happened already, when this gets matched up with data on the dark web from other hacks, I could imagine how this could lead to new kinds of personal intimidation and terrorism by foreign influences even in the US, not just by the government.   Hasn’t really happened yet.
  
Look at the Fibbies page on Media Monitoring Services.  

Friday, April 06, 2018

The bad karma of free content -- and how it feeds the new cultural wars begging for authoritarianism


Once again, we ponder one big problem with user-generated content from gratuitous speakers like me – karma. That is, using a "free" (more or less) platform that could not exist if it couldn't allow bad behavior of others.  This gets down to personalizing the question, "Am I my brother's keeper?"  We could wind up telling speakers that they must be so, or else they become thieves.  Yup, it sounds a little Commie, even if it comes from the far right as well as the left. 
  
No, I don’t spam, recruit people, or gather user information, or traffic anything, display porn (I do get into PG-13 territory), and so on.  I also don’t support specific candidates in elections – which is both good and bad.  I am not very partisan and not tribalist.  I try to use reputable sources.    On something like the gun debate, there really are two sides – and the opposing camps serve different positions in life on this planet. So what I try to do is remind visitors of the nuances on so many interrelated policy questions (even if this all started with “don’t ask don’t tell” two decades ago).  With some issues, like DADT, my approach has been influential. With some others, maybe I'm "in the way". 
  
And I am able to do this without joining anyone else’s “intersectional” movement and screaming (with "solidarity") in a crowd about oppression (particularly without carrying somebody else's picket signs).  In my case, such claims for myself would make no real sense.
  
What’s bad is that I depend on an Internet (including social media and Facebook) infrastructure that needs a business model to pay its way, and that business model could not exist without some actors doing bad things.  So we’ve needed Section 230 to protect providers from downstream liability, because there is no way from them to tell bad from innocuous prospectively.  Furthermore, the most recent news stories on the Facebook scandal suggest that these platforms could not pay their own way without the “bad actors” using them and driving ad revenue (and share prices).
  
  
I remember Donald Trump’s December 8, 2015 (post here) proposals to “shut down those tubes” and even Hillary Clinton hinted at that, to curb the gratuitousness that has led, for example, to so much cyberbullying, trafficking (FOSTA), terrorist recruiting and instructions on how to build weapons (which are out there in print anyway). 
  
Then Trump (despite his distrust of computers and now his hatred of Amazon) went the other direction, using Twitter to announce his policies (like the half-transgender ban in the military) and his grievance against his own version of the improperly privileged. He seems to have much more gripe against established media companies (except Fox, and now OANN and Sinclair which are getting more prominent)  than individualistic and perhaps libertarian oriented speakers (me, even Milo). 
Could the “bad karma” lead to some kind of edict shutting us all down?  We can imagine that war happens first – a nuclear blast (or more than one) somewhere from North Korea, or an EMP level one attack that might shut down tech companies unless they had caged their data (there is some evidence they are doing just that, not talked about much).
  
As a Milo-Dangerous thought experiment, we can imagine that no free content is allowed:  everything that is available to the entire public has to pay its own way.  Free blogs (like this one) are gone.  Of course, you can say, you can pay for hosting.  Or you can pay to self-publish a book.  But even that goes.  Anything out there has to meet a real “need” or “want” that a consumer will pay for. That could mean that even a self-paid book has to sell minimum volume or it is speaker is shuttered, maybe forever.  Then at least you solve the karma problem of the business models.  Milo has recently made most of his content paid subscription, which I first thought was curious, but could that be his point?  I can think of a lot of obvious problems with this idea – starting with porn (when Congress has just cracked down on supporting sex trafficking and prostitution, remember?)

Of course, spontaneous user-generated content backs up a free press (the Fifth Estate backs up the Fourth) and is a check on authoritarianism.  It’s a little harder for potential dictators to get away with things.  On the other hand, hyperindividualsim can, if challenged, lead to a reaction that causes a desire for authoritarianism and “rules consistency” to bubble up from the bottom, and lead back to fascism.
  
I do wonder if we’re entering a war on introverts and sophists (certainly schizoids and aspie-types -- and all the Rosenfels "subjective feminines").  The simple idea of “personal responsibility” that libertarians articulated so well late in the Clinton years (until 9/11) has become layered.  Now, your immediate community as a whole means something, which is a way to giving less competitive (and more tribally minded) people in that community more meaning.  That means everyone has to lean to take care of their own first – including the childless  - everybody must "belong" to some group.  [Right off the bat, I wonder about the incredible generosity to strangers I do see online, like the willingness to undergo intrusive organ donation procedures, an idea unthinkable in the more closed culture that I grew up in.] You see that in the talk of “demographic winter” and in authoritarian leaders’ (like Putin’s) idea that the gay community is discouraging other people from procreating.  We may well be heading toward a culture where you are not heard from until you have your own skin in the game – like having your own children first.  
  
What seems so shocking is that a lot of people around the world want this kind of culture back. But it  isn’t as surprising as we think.

Wednesday, April 04, 2018

Facebook drastically narrows its privacy policies for app developers, after count of compromised accounts rises to 87000


Facebook has announced major changes in its policies with regard to the ability of developer apps to access user and especially Friends’ data, with its own news story link here.  Techcrunch has a nice summary.  CNBC reports here
  
I have often been queasy about announcing (with checkin) that I would attend certain events, particularly when I was living alone in a detached house.  There are situations where I would prefer not for it to be known in advance on who will come.
  
Facebook has raised the estimate of the number of people whose personal information was impropertly shared with Cambridge up to 87 million, from 50 million.
  

The new changes will perhaps make account recovery more difficult, especially in countries with certain personal naming language conventions (mostly in Asia).




Facebook has also announced some restrictions on its own subsidiary, Instagram (the marketeer's dream),  My own doesn't work in Chrome on Windows right now. 
  
Mark Zuckerberg will testify before Congress on Wednesday, April 11, 2018.  He has announced certain changes requiring verification of some users already, story