Summary of “The end of the internet startup”

Some critics say, they’ve gotten better at controlling and locking down key parts of the internet’s infrastructure, closing off paths that early internet companies used to reach a mass market.
Most important, Google bought a little-known mobile software company called Android in 2005, laying the foundation for Google’s eventual dominance of smartphone operating systems.
If these companies had remained independent, they easily could have emerged as major competitors to Google and Facebook.
Yelp CEO Jeremy Stoppelman rebuffed acquisition offers from Google and Yahoo, taking the company public in 2012.
“At one point,” writes Businessweek’s Brad Stone, “Quidsi executives took what they knew about shipping rates, factored in Procter & Gamble’s wholesale prices, and calculated that Amazon was on track to lose $100 million over three months in the diaper category alone.” As a venture-backed startup, Quidsi couldn’t sustain those kinds of losses, so the company wound up selling to Amazon in 2010.
“Mark Zuckerberg had a huge advantage with Facebook because the pressure that normal people have of building a company was replaced by the lightness of him just playing around with ideas,” said Mike Maples, an investor at the firm Floodgate.
In the 1980s, great companies like Microsoft, Adobe, and Intuit were founded to make software for PCs. Those companies still make plenty of money – just like Intel does – but there isn’t a lot of room for desktop PC software startups today.
There are only so many things you can do with a web browser or a smartphone, and maybe companies like Google, Facebook, and Snap have already locked down the most important markets.

The orginal article.

Summary of “The Facebook Algorithm Mom Problem”

Anecdotally most in social media have long known that doing this type of workflow causes your content to be treated like a second class citizen, particularly on Facebook which greatly prefers that users post to it manually or using one of its own apps rather than via API. This means that the Facebook algorithm that decides how big an audience a piece of content receives, dings posts which aren’t posted manually within their system.
Simply put, if you don’t post it manually within Facebook, not as many people are going to see it.
My mom, who seems to be on Facebook 24/7, immediately clicks “Like” on the post.
The Facebook algorithm immediately thinks that because my mom liked it, it must be a family related piece of content-even if it’s obviously about theoretical math, a subject in which my mom has no interest or knowledge.
Now Facebook’s algorithm has created a self-fulfilling prophesy and further narrows the audience of my post.
As a result, my post gets no further exposure on Facebook other than perhaps five people-the circle of family that overlaps in all three of our social graphs.
I can post about arcane areas like Lie algebras or statistical thermodynamics, and my mom, because she’s my mom, will like all of it-whether or not she understands what I’m talking about or not.
The problem is: Facebook, despite the fact that they know she’s my mom, doesn’t take this fact into account in their algorithm.

The orginal article.

Summary of “Facebook is getting ready to test paid subscriptions with publications”

These options would seem to accommodate metered publications including Facebook skeptic The New York Times; and The Wall Street Journal and The Economist, which make a certain selection of articles free and put the rest behind a paywall.
One publisher said Facebook is leaning toward letting them be handled through the mobile web, which would eliminate a layer of complexity and the app stores’ cut and would let publishers have more control over pricing and data.
In the end, the question for publishers like the Times and Journal that have been cool on Instant Articles is whether there’s an advantage to Facebook’s proposal of letting them sell subs through Instant Articles instead of posting to Facebook the old-fashioned away and letting people subscribe by traveling to the publisher’s own site.
Facebook aims to test them with a small group of publishers at the end of the year and expand them to others in 2018.
The movement on subscription sales comes as Facebook has been trying harder to win over publishers that are getting increasingly frustrated over their lack of ability to monetize their content on the platform and are also getting in some cases what they feel is better treatment from other platform giants, Google and Apple News.
As part of its charm offensive, Facebook in January launched the Facebook Journalism Project, which has involved working with publishers to understand what they want to get from the app.
Publishers have been seeking ways to sell subscriptions since Facebook rolled out Instant Articles in 2015 and have been been clamoring for more ability to monetize their content on the platform through advertising as Google and Facebook have swallowed up more and more of the digital ad pie.
“We are in early talks with several news publishers about how we might better support subscription business models on Facebook,” Campbell Brown, head of news partnerships, said via a Facebook spokesperson.

The orginal article.

Summary of “Facebook can track your browsing even after you’ve logged out, judge says”

A judge has dismissed a lawsuit accusing Facebook of tracking users’ web browsing activity even after they logged out of the social networking site.
The plaintiffs alleged that Facebook used the “Like” buttons found on other websites to track which sites they visited, meaning that the Menlo Park, California-headquartered company could build up detailed records of their browsing history.
Clicking on the Facebook “Like” button on a third party website – for example, theguardian.com – allows people to share pieces of content to Facebook without having to copy and paste the link into a status update on the social network.
Australian internet security blogger Nik Cubrilovic first discovered that Facebook was apparently tracking users’ web browsing after they logged off in 2011.
Responding to Cubrilovic, Facebook engineer Gregg Stefancik confirmed that Facebook has cookies that persist after log-out as a safety measure but that the company does not use the cookies to track users or sell personal information to third parties.
In 2014 Facebook started using web browsing data for delivering targeted “Interest-based” advertising – which explains why you see ads for products you have already been looking at online appear in your Facebook feed.
To address privacy concerns, Facebook introduced a way for users to opt out of this type of advertising targeting from within user settings.
“We are pleased with the court’s ruling,” said a Facebook spokeswoman.

The orginal article.

Summary of “There’s No Money in Internet Culture”

As Verizon completed its acquisition, a number of Tumblr employees, as well as those at other Verizon-owned properties, like the Huffington Post, were laid off.
The site is enormously popular among the coveted youth crowd – that’s partly why then-CEO Marissa Mayer paid $1 billion for the property in 2013 – but despite a user base near the size of Instagram’s, Tumblr never quite figured out how to make money at the level Facebook has led managers and shareholders to expect.
Even once the timeline became open to advertising, it was tough to find clients willing to brave the sometimes-porny waters of the Tumblr Dashboard.
Looked at from a bottom-line perspective, Tumblr is an also-ran like its parent company – a once-hot start-up that has eased into tech-industry irrelevance.
Looked at from another angle Tumblr is among the most important sites online – a central hub of what is nebulously known as “Internet culture.” Most recently, the site gave us Dat Boi, the unicycling frog, but Tumblr’s most famous legacy is probably the reaction GIF, which was popularized by Tumblr accounts like What Should We Call Me. Tumblr’s reblog structure, which created lengthy, publicly shared conversations between strangers, also helped popularize the concept of the Discourse, the internetwide conversation happening all at once.
There are many reasons for this – for one thing, it’s tough to sell a high-quality ad experience to executives at Coca-Cola when you first have to explain what a meme is and why it’s “Viral.” On top of all that, there are reams of porn, hate speech, copyright infringement, and more porn floating around on these platforms, easily accidentally placed adjacent to a company’s studiously inoffensive ad. Maybe more importantly, Tumblr and Vine and the like never had data-mining operations as sophisticated as, say, Facebook.
Which means Tumblr has to hope for patience and kindness from Verizon while it seeks a way to make money.
The overall stumbles of building centralized hubs of internet culture mean that, going forward, content might soon be consumed not by one large audience on a single platform, but by thousands of smaller audiences across a variety of online spaces.

The orginal article.

Summary of “Facebook’s Secret Censorship Rules Protect White Men from Hate Speech But Not Black Children”

A trove of internal documents reviewed by ProPublica sheds new light on the secret guidelines that Facebook’s censors use to distinguish between hate speech and legitimate political expression.
There is a ban against pictures of Pepe the Frog, a cartoon character often used by “Alt-right” white supremacists to perpetrate racist memes, but swastikas are allowed under a rule that permits the “Display [of] hate symbols for political messaging.” In the documents examined by ProPublica, which are used to train content reviewers, this rule is illustrated with a picture of Facebook founder Mark Zuckerberg that has been manipulated to apply a swastika to his sleeve.
Facebook’s director of policy for the region, Thomas Myrup Kristensen, acknowledged at the time that it “Found a small number of accounts where we had incorrectly removed content. In each case, this was due to language that appeared to be hate speech but was being used in an ironic way. In these cases, we have restored the content.”
In May, she deplored the increasingly common Facebook censorship of black activists in an article for Medium titled “Mark Zuckerberg Hates Black People.”
After a year of negotiations, Facebook, Microsoft, Twitter and YouTube agreed to the European Union’s hate speech code of conduct, which commits them to review and remove the majority of valid complaints about illegal content within 24 hours and to be audited by European regulators.
Now the German government is considering legislation that would allow social networks such as Facebook to be fined up to 50 million euros if they don’t remove hate speech and fake news quickly enough.
Facebook recently posted an article assuring German lawmakers that it is deleting about 15,000 hate speech posts a month.
Worldwide, over the last two months, Facebook deleted about 66,000 hate speech posts per week, vice president Richard Allan said in a statement Tuesday on the company’s site.

The orginal article.

Summary of “Mark Zuckerberg explains why he just changed Facebook’s mission”

Zuckerberg believes he has just the tool for the job: Facebook Groups, which are now used by a billion people.
Three hundred Facebook Group administrators from across the country are attending the two-day event to hear speeches from Zuckerberg and other executives, and attend panels on topics like conflict resolution.
On Facebook: ‘I think we’re doing OK’. The new emphasis on Groups is the culmination of months of public appearances and posts by Zuckerberg stressing the importance of community.
It’s part of what pushed Zuckerberg to reexamine Facebook’s mission, starting with a 5,726 word post on “Building Global Community” in February.
The way Zuckerberg describes it, being active in Groups can be a political act, a grassroots way to address the ills of the world.
Most groups are casual – Zuckerberg himself is in puli groups, where people share cute photos of dogs like his own famous pet, Beast.
Only 100 million Facebook users are in what Zuckerberg calls “Meaningful” groups – the kinds that become part of your support structure.
Facebook has started using artificial intelligence to identify users or groups trying to engage in terrorist recruiting, and Zuckerberg says they’ll do more over time as the artificial intelligence gets better.

The orginal article.

Summary of “Media Companies Are Getting Sick of Facebook”

When Facebook Inc. wants to try something new, one of its first calls is to CNN. It was a key partner when Facebook introduced its news-reading app, Paper, in 2014.
“For them, these are experiments, but for the media companies looking to partner with ¬≠significant commitments, it gets to be a bit of whiplash.” Morse says the financial compensation Facebook offers isn’t enough to convince him that working directly with the social network will be worthwhile in the long term.
The companies rely heavily on Facebook to boost their audiences, especially on mobile devices.
Outlets have started to pull back from Facebook partnerships concerned that they’re putting more into the deals than they’re getting out of them.
Still, Simo says media com¬≠panies shouldn’t expect to get the upfront payment some are calling for, partly because Facebook rewards viewership and partly because viewership is its own reward.
While Facebook is scouting deals for shows and projects that can compete with YouTube this summer, for now it needs other companies to produce videos its 1.9 billion users will want to watch.
On June 8, Facebook announced it would allow publishers to put more ads in Instant Articles and said it’s been paying $1 million in total ad revenue daily to the companies using the service.
Ben Lerer, CEO of online video machine Group Nine Media, says that while he’s not satisfied by his deals with Facebook, he’s optimistic they’ll improve.

The orginal article.

Summary of “Revealed: Facebook exposed identities of moderators to suspected terrorists”

Facebook put the safety of its content moderators at risk after inadvertently exposing their personal details to suspected terrorist users of the social network, the Guardian has learned.
A bug in the software, discovered late last year, resulted in the personal profiles of content moderators automatically appearing as notifications in the activity log of the Facebook groups, whose administrators were removed from the platform for breaching the terms of service.
The personal details of Facebook moderators were then viewable to the remaining admins of the group.
Facebook moderators like him first suspected there was a problem when they started receiving friend requests from people affiliated with the terrorist organizations they were scrutinizing.
An urgent investigation by Facebook’s security team established that personal profiles belonging to content moderators had been exposed.
Facebook then discovered that the personal Facebook profiles of its moderators had been automatically appearing in the activity logs of the groups they were shutting down.
In one exchange, before the Facebook investigation was complete, D’Souza sought to reassure the moderators that there was “a good chance” any suspected terrorists notified about their identity would fail to connect the dots.
The Guardian recently revealed the secret rules and guidelines Facebook uses to train moderators to police its vast network of almost two billion users, including 100 internal training manuals, spreadsheets and flowcharts.

The orginal article.

Summary of “An Artificial Intelligence Developed Its Own Non-Human Language”

A buried line in a new Facebook report about chatbots’ conversations with one another offers a remarkable glimpse at the future of language.
In the report, researchers at the Facebook Artificial Intelligence Research lab describe using machine learning to train their “Dialog agents” to negotiate.
At one point, the researchers write, they had to tweak one of their models because otherwise the bot-to-bot conversation “Led to divergence from human language as the agents developed their own language for negotiating.” They had to use what’s called a fixed supervised model instead. In other words, the model that allowed two bots to have a conversation-and use machine learning to constantly iterate strategies for that conversation along the way-led to those bots communicating in their own non-human language.
The detail about language is, as one tech entrepreneur put it, a mind-boggling “Sign of what’s to come.”
They do demonstrate how machines are redefining people’s understanding of so many realms once believed to be exclusively human-like language.
Already, there’s a good deal of guesswork involved in machine learning research, which often involves feeding a neural net a huge pile of data then examining the output to try to understand how the machine thinks.
The fact that machines will make up their own non-human ways of conversing is an astonishing reminder of just how little we know, even when people are the ones designing these systems.
“There remains much potential for future work,” Facebook’s researchers wrote in their paper, “Particularly in exploring other reasoning strategies, and in improving the diversity of utterances without diverging from human language.”

The orginal article.