OLDaily

Why AI Will Never Replace Managers
Frank Martela, Jukka Luoma, Harvard Business Review, 2021/09/07

I think this another one of those 'technology will never...' posts that gets proven wrong, and even laughed at, within a few decades. This one is particularly egregious because it misunderstands what AI can do, misunderstands the contribution, and misunderstands even what managers do. Most management activity is in the area of compliance, not goal setting. And this is really easy for a machine to do (so easy, in fact, there is a substantial anti-surveillance sentiment). And while the authors say "AI is simply a variety of prediction algorithms," so is human cognition. There's not some special skill like 'reframing' that is distinct from all other cognition. So as much as the authors tout 'thinking slowly' as something only humans (and in particular, managers) can do, I think we'll find very quickly that AI can create and optimize various scenarios a lot more efficiently than humans.

Web: [Direct Link] [This Post]

The EdX Aftermath
Michael Feldstein, e-Literate, 2021/09/07

Michael Feldstein gets straight to the point of the major impact of the sale of OpenEdX: "I’m far more interested in the ramifications of the sale on public-good collaboration in higher education," he writes. "I think Harvard and MIT have done significant damage to the landscape and to themselves.... It’s the hypocrisy. The creation of EdX was positioned as a direct response to those dirty, grubby capitalists from Stanford who were going to privatize higher education with Coursera. EdX, in contrast, was to be a non-profit."

Web: [Direct Link] [This Post]

Installing Owncast on Reclaim Cloud
Jim Groom, bavatuesdays, 2021/09/07

If you want to offer streaming video but you don't want to use YouTube, Owncast offers an open source alternative. Jim Groom describes Owncast with a video and text in this article, and then (to make this useful for everyone, provides a video and instructions showing how to install it on Reclaim Cloud.

Web: [Direct Link] [This Post]

Popper and the Propensity Interpretation of Probability
Charles H. Pence, PhilSci Archive, 2021/09/07

I've often used three interpretations of probability - frequency, logical, and subjective - to illustrate the idea that the same logical or mathematical foundation can be applied to the world in different ways. To this I can now add Popper's 'propensity interpretation'. "On this view, objects in the world have dispositions or tendencies to bring about or realize certain kinds of future possible outcomes." It's not merely based on observation, as is the frequency interpretation, and not based on ontology, as in the logical interpretation. It has been around for decades, of course, but it was new to me. Image: Kidney International.

Web: [Direct Link] [This Post]

MyNotes: Exploring Humanism, Chapters 1-2
Miguel Guhlin, Around the Corner, 2021/09/07

Miguel Guhlin shares his summary of Greg Epstein's book, Good Without God, and I thought I'd include it here as complexion and nuance for my upcoming MOOC. It follows a path well trodden by people like Kai Nielson and J.L. Mackie. I appreciated Guhlin's honest and forthright consideration of the work. There are three posts, covering chapters one and two, chapter three, and chapters four and five. It considers the questions not only of whether humanists can be good, but even questions like "what are we striving for", including such things as human dignity and compassion.

Web: [Direct Link] [This Post]

How To Read AI Papers
Kirill Demochkin, Casual GAN Papers, 2021/09/07

Although this article is specific to artificial intelligence (AI) papers, it has a broader application, and I find many of the principles apply to papers in other domains, including online learning. A paper is not like a novel you read from beginning to end. Different types of papers are read differently. For example, I will read a formal academic paper starting with the abstract, then decide to look at the literature review (if the area is new to me) or the discussion (if it's not). If the discussion offers interesting results, I'll look at the methodology. For a stream-of-consciousness blog post, by contrast, I'll often read from the bottom up, so I get the author's main point first, then work back through the chain of reasoning that brought them to it. The main point here is: know what you're looking for, and have a strategy for finding it in the paper (if it's there). Image: Flame-In-Nerf.

Web: [Direct Link] [This Post]

Accesible and Universal Design for Learning in OER
Josie Gray, BCcampus, 2021/09/07

This is a set of speaking notes for a talk on designing for accessibility in learning materials. The notes stand up quite well on their own, offering a detailed and useful overview of the subject, looking at the four major dimensions of accessibility: perceivable, operable, understandable, and robust.  What I like also is the statement near the end that "Students may face all sort of barriers even if they don’t have a diagnosed disability, and there is not really such thing as an 'average student.'" Quite right. "The reality is, there is no such thing as a normal or average student. Students vary greatly in their interests, family situation, culture, background, experience, strengths, and weaknesses. And all students benefit when educational materials are designed to be accessible and inclusive." There's a lot more in addition; this is a thoughtful and complete presentation that I can easily recommend.

Web: [Direct Link] [This Post]