OLDaily

Google to restrict modern ad blocking Chrome extensions to enterprise users
Kyle Bradshaw, 9to5Google, 2019/06/03

Here we have a case where a media platform has become sufficiently dominant that it is prepared to enforce the requirement that users view ads (and get tracked, and infected with malware, and the rest of it). It's just this sort of case that induces me to use Firefox (which I have pretty consistently over the years) instead of a commercial product. As uBlock Origin developer Raymond Hill says, "Google's primary business is incompatible with unimpeded content blocking. Now that Google Chrome product has achieve high market share, the content blocking concerns as stated in its 10-K filing are being tackled." More: BGR ("Google quietly ruined Chrome..."), Vice ("Google struggles to justify..."), CNet ("Google holds firm..."), Forbes ("Google Just Gave 2 Billion Chrome Users A Reason To Switch To Firefox"). If I can't prevent a browser from loading unwanted and unsafe content, there's no way I'm running it on my desktop. Image: DazeInfo.

Web: [Direct Link] [This Post]

What Works to Reduce Inequalities in Higher Education?
Koen Geven, Estelle Herbaut, World Bank, 2019/06/03

Alex Usher points to this World Bank report (89 page PDF) subtitled "A Systematic Review of the (Quasi‐)Experimental Literature on Outreach and Financial Aid." The authors "are exclusively concerned with outcomes of disadvantaged students" and focuses "on both enrollment in and completion of higher education." After a survey of the barriers faced by disadvantaged students, the report looks at specific interventions (their word, not mine) and assesses their success rate as revealed in the literature (the volume of which is in some cases unimpressive).

The results? "Outreach interventions targeted at students in high school or recent graduates seem to be a relatively cost‐effective tool to address inequalities in access to higher education, as long as the interventions go beyond providing general information about higher education." Support needs to be continued during the higher education period as well. Meanwhile, while "an early commitment of (needs-based) aid, while students are still in high school, leads to much larger impact on higher education access," by contrast, "merit‐based aid based only on academic results, without any assessment of students’ financial needs, seems to have no effect." Overall, "Interventions that combine early financial aid and outreach activities... seem to lead to large increases in enrolment rates, more consistently than either outreach or financial aid alone."

Web: [Direct Link] [This Post]

Post-18 review of education and funding: independent panel report
Jeremy Augar, Gov.UK, 2019/06/03

This report, known as the 'Augar Report' (216 page PDF), has commanded univocal attention from the British educational media over the last few days. Reaction has been mixed. Many critics agree with the assessment from the Financial Times to the effect that "the Augar panel’s wider proposals on reforms to student finance are regressive.... Shockingly, Augar proposes — explicitly — that the changes come at the expense of lower and middle-earning graduates, who would pay more." As Justine Greeing opines, it "represents much that has gone wrong in British politics." Hard to disagree. Much more on Augar from University World News, TES, The Conversation, The Guardian, HESA (Alex Usher), WonkHE (which has no fewer than 16 articles on Augar), Schools Week,

Web: [Direct Link] [This Post]

Goodhart’s Law: Are Academic Metrics Being Gamed?
Michael Fire, The Gradient, 2019/06/03

The answer to the question in the title is "Yes, yes, of course they are." The gaming probably goes well beyond the trends observed in this article. Trends include these: "over time, papers became shorter while other features, such as titles, abstracts, and author lists, became longer"; and "a sharp increase in the number of new authors (who) are publishing at a much faster rate given their career age"; and "a drastic increase in the number of ranked journals, with several hundred new ranked journals appearing each year"; and "different domains had widely ranging properties. Even subfields of the same domain had surprisingly different average numbers of citations." Via Reddit.

Web: [Direct Link] [This Post]

Public Reason Liberalism
Jerry Gaus, 2019/06/03

In my E-Learning 3.0 course I advanced the idea of "community as consensus". This article (22 page PDF) allows us to look at some of the historical roots of that idea, as found in social contract theory (in, eg., Hobbes, Locke and Rawls), of 'public reason', that is, when people disagree "a cooperative and just social life requires that they abandon their private judgment about their claims and submit to the public reason of impartial justice." This in turn allows me to consider how exactly how (or whether) modern consensus algorithms are a departure from (and maybe an advance on?) traditional social contract theory.

Web: [Direct Link] [This Post]

Cake or death: AMP and the worrying power dynamics of the web
Andrew Betts, 2019/06/03

Andrew Betts lays out the argument against Accelerated Mobile Pages (AMP). A lot of the discussion is pretty technical, but they boil down to the assertion (which is well-founded) that AMP essentially turn the web into a private portal for Google, and that Google is not managing this stewardship responsibly. As Betts says, "If Google was my doctor, they’d be currently explaining to my family that although the experiment they tried did sadly kill me, they got a ton of useful data from it, and they think they can definitely work on fixing that bug in the next version of the experiment." The article also links to a really interesting set of resources worth exploring, including the Portals specification, feature policy, Newsguard, and content passes (which "would require publishers of paywalled content to declare the paywall in the metadata of the article" instead of trying to trick people into clicking). Via Aaron Davis.

Web: [Direct Link] [This Post]

Our AI Overlords Are Already Here, They Likely Employ You
Ton Zijlstra, Interdependent Thoughts, 2019/06/03

Short post pointing to a reprint of an article by Jeremy Lent from late 2017 arguing that corporations are artificial intelligences. This framing of the corporation allows us to talk about the ethics of AI in a different way, writes Ton Zijlstra. "It refocuses us on the fact that organisational structures are tools. When those tools get bigger than us, they stop serving us," he says. And second, it puts questions about ethics well into the future, "and not paying attention to how those same ethical issues play out in your current context." Imagine subjecting corporations to the same sort of ethical scrutiny we do to AI. Privacy, autonomy, choice - all of these are impacted by corporation in ways we would deem unacceptable for an AI.

Web: [Direct Link] [This Post]

‘Robots’ Are Not 'Coming for Your Job'—Management Is
Brian Merchant, Gizmodo, 2019/06/03

The Davos set likes to talk about the "4th industrial revolution", which sounds great, but serves to mask the core issue of the previous three, in which the benefits earned by increasing industrialization benefited only a few, while the many who were displaced had to fight to retain even a small stake in society (and the same is true of the agricultural revolutions of earlier years, as the history of people like John of Gaunt will tell you). What this article stresses is that industrialization - this time in the form of robots - is something people do to other people. "The CEOs who see an opportunity to reap greater profits in machines that will make back their investment in three point seven years and send the savings upstream—they’re the ones coming for your job."

Web: [Direct Link] [This Post]