Stephen Downes

This is "a modern fully open-source textbook exploring why and how deep neural networks learn compact and information-dense representations of high-dimensional real-world data." It's a daunting read - and to be honest, I gave the first chapter a careful read but have only skimmed the rest (though it's now on my virtual bookshelf for future reference). I think people who are serious about understanding how we learn and how machines emulate this should at the very least read that first chapter. It contains what is to my understanding an accurate history of artificial intelligence (including the important early work in computation, cybernetics, and information theory). The aim of the book is to give us a framework that explains how neural networks work to generate intelligence. They authors write, "intelligence is not about collecting all informative data in advance and training a model to memorize their (low-dimensional) distribution. Instead, it requires computational mechanisms that can continuously improve current knowledge of the distribution and acquire new information when needed. A fundamental characteristic of any intelligent being or system - whether an animal, a human, a civilization, or the entire scientific community - is the ability to continuously improve or gain new information or knowledge on its own."

Wendy Palmer comments on Accredible's State of Credentialing report which argues, among other things, that "Credentialing programs have entered an accountability era. As expectations rise, organizations are increasingly expected to demonstrate the impact of their programs - not just issue credentials." She picks up on that point to explore "how we measure the value of digital credentials, and whether the measures we're using really tell us what we need to know about their quality and impact." She asks, "are we conflating the quality of the learning experience with the quality of the credential that represents it?" After all, "they aren't quite the same thing. A course might be exceptionally well designed, delivered and assessed. It might meet every institutional quality standard. Yet the credential issued at the end often simply says that someone successfully completed the course. What does that communicate to an employer." Via Doug Belshaw. 

Services around the world that depend on PHP (that would include things like WordPress and Pixelfed) are either offline or being repaired today as a result of a new and critical PHP vulnerability. Here's the official announcement. I have one website running on PHP, my leftish.media blog. I checked and there isn't a patched PHP Docker container available yet, so I set a reminder to check tomorrow. Anyhow, this page has information about the vulnerability and, more importantly, what you need to recover from it.

The Decentralized Identity Foundation (DIF) universal distributed ID (DID) resolver has been shot down, with users now being directed to a second tier of providers, including especially GoDiddy, who created the DIF's reference implementation, and thisDID, by GoPlausible. "Both deployments are deployed and funded by DIF members, who maintain the respective reference implementations they are based on through DIF." This might be thought of as a maturation of the specification, though of course there are risks that it will be commercialized and enclosed. See also the DIF Identifiers & Discovery Working Group.

Doug Belshaw links to and discusses the idea to the 'scenius'. The linked article notes, "Culture isn't created by outliers. Most new movements and leaps forward in culture usually come from groups of creatives encouraging each other; competing with and arguing with each other; reacting for or against each others' work." Belshaw comments, "there's a new article out by Drs Helen DeWaard, Verena Roberts, Leo Havemann, and Gabi Witthaus entitled Disputatio in the Digital Agora: Opening the Doctorate Through Dialogue. I wonder what would have happened had my own doctoral studies eveolved in the open, among a community of interested people. That wasn't an option back then, and my pursuit of connectionist knowledge in a philosophical sense was a solitary struggle that was ultimately unsuccessful in landing me the title and the degree.

"If personal AI tutors seem too good to be true," writes  Mia Travers-Hayward, "that's because they are. This vision is built on a lie. Tech won't save our school systems. AI tutors can't substitute the care of a human teacher." I have two things I need to say about this (you can read the argument and supporting evidence for yourself). First, it is way too early to be talking about what AI can or cannot do in education. Second, the whole concept of 'AI tutors' is based on some sort of direct instruction fantasy model, and this is not how education in the future will play out. Will people still need care in the future? Sure. Will they need to get it from teachers? Debatable.