OLDaily

25 Ed Tech Predictions for 2021
Dian Schaffhauser, Campus Technology, 2021/01/05

I won't try summarizing these, except to say that it's another one of those articles that highlights short predictions by 25 different people (most of which are vacuous, but some of which contain a nugget of insight). There's a lot of overlap and repeats from the selection I ran yesterday (oh oh chatbots I'm so excited).

Web: [Direct Link] [This Post]

The next normal arrives: Trends that will define 2021—and beyond
McKinsey, 2021/01/05

I don't think the new normal or the next normal is arriving just yet, but we can see ahead to think that it might be coming in a few months. This is McKinsey's take on what that looks like. You'll need patience with this page - there's a lot of scrolling for bits of text (or you can use the Tranquility Reader Firefox extension, like I did). Anyhow, McKinsey expects a consumer rebound, consumer travel (but not so much business travel), a wave of entrepreneurs, an acceleration of the "Fourth Industrial Revolution" (ie., artificial intelligence), changes in shopping behaviour (ie., to online+delivery), more work from home, biopharma, environmental issues, and way down at the bottom (and probably most significant) government attempts to tackle debt.

Web: [Direct Link] [This Post]

A Post-Covid Case for Classroom Cameras
Michael J. Petrilli, Education Next, 2021/01/05

I think it's pretty certain that post-Covid classrooms will employ a lot more technology than they used to. Even if nothing else changes, digital content is set to become mainstream. But what about all technology? It won't be equally welcomed. Here, for example, is a call to record in-class sessions. Personally, I would think it's fine for most cases, but the same people who object to surveillance technology in learning will likely object to class recording, and for similar reasons. Michael Petrilli says, "no doubt some teachers won’t be thrilled to know that their administrators could also be watching them—and evaluating their performance—any time they like." The same applies to students, perhaps even more so, especially is analytics are applied to the live feed.

Web: [Direct Link] [This Post]

Why I’m lukewarm on graph neural networks
Matt Ranger, Single Lunch, 2021/01/05

I'm posting this here mostly for the image, and you don't even see the image in the Reddit version unless you click on a specific link. the image makes clear in a way no number of would could the relation between a network (that is, the graph, that is, the nodes connected by connections) and a matrix (that is, a set of numbers arranged in rows and columns, which also denote the connections). Most neural net software is based on performing mathematical operations on the matrix, but that's hard to visualize when looking at patterns in the network. The rest of the post is pretty interesting, but the whole point is really made with the diagram.

Web: [Direct Link] [This Post]