“Canary staff given access to [previously] ‘restricted systems’ found evidence of ‘gross inequalities and gross mismanagement that was rampant in the company’…there had been failings in the management of the site’s membership database and that overall there was a ‘culture of hierarchy and narcissism.'”

“While all White House press reporters attracted their trolls, with Haberman there was an order-of-magnitude difference. A sizable faction of very online folks clearly believe that Haberman was not covering Trump, but rather covering for Trump…First, in her Twitter feed, Haberman gave more weight to stories that offered a more favorable view of Trump’s political machinations, which annoys folks who despise Trump. Second, never underestimate plain-old sexism: as a woman, Haberman gets far more vitriol than her male counterparts.”

“In retrospect there were probably better ways to try to convey the danger posed by Trump than to serially mock him on Twitter, reinforcing the editorial decisions that treated his tantrums but not his actions as the news, even while exacerbating the polarization between those who identified with Trump’s tantrums and those who with their fancy PhDs knew better….[Haberman’s] stories, individually and as a corpus, revealed Trump to be a skilled bully. But those stories of Trump’s bullying commanded our attention, just like his reality TV show did, and reassured him that continued bullying would continue to dominate press coverage.”

“I don’t think [Americans focused more on scandalous White House anecdotes than Trump’s actions] because reporters like Haberman were engaging in a pattern of distraction. I think it happened because Americans prefer to read about easily digestible, scandalous stories like that than longform deep dives into policy. If Wheeler and others want to argue that Haberman et al’s reporting does not have a lot of nutritional value, that is an utterly defensible position. But I don’t think that this coverage crowded out better coverage. It was all available to read. It’s just that most Americans, when they read about politics at all, will go for the quick sugar high.”

“…the ‘experiment’ is meant to let users elevate their chat messages for a specific time using a one-time fee. The fees are presented in five different tiers ranging from 30 seconds to two and a half minutes, with fees ranging from $5 to $100.”

“Botmasters have created a Kafkaesque system where companies are paying huge sums to show their ads to bots. And everyone is fine with this….'[A problem] that no one talks about, no one writes about, everyone thinks it’s someone else’s problem.'”

“As a title, Glitz may ring more gossip rag than hard-hitting media, but the teaser stories it already has posted online are detailed financial and business articles and, according to Indigo, its target audience ranges from luxury consultants to media outlets. On its website, Glitz calls itself ‘the first investigative publication dedicated to the global luxury sector,’ a market that, according to a recent report by Bain & Company, could surpass $300 billion this year.”

“Such participatory journalism has no equal in the news business. War reporters usually do not place themselves in the midst of combat, and police reporters typically do not do ‘standups’ in the middle of a shootout. A reporter covering a fire keeps a safe distance.”

“Outsourcing the heavy lifting of data collection and model training to non-commercial entities allows corporations to avoid accountability and potential legal liability. It’s currently unclear if training deep learning models on copyrighted material is a form of infringement, but it’s a harder case to make if the data was collected and trained in a non-commercial setting.”

“This means that its use as a reference in Wikipedia articles will not be permitted for ‘exceptional claims’ that require heightened scrutiny, but that its reliability will be evaluated on a case-by-case basis for other claims.”