Please enjoy my most recent photos, a set from the Quebec resort town of Mont Tremblant, taken while we visited for my workshop on Satuday. It's Andrea's new favourite place and I would also return any time. Click on the photo, headline, or here to see the set.
Rice University's OpenStax, which once upon a time was known as a free textbook site called Connexions, has launched Tutor Beta, an online learning platform that will make three courses -- biology, physics and sociology -- available this fall. I'm thinking that the textbook project felt it was missing out on the MOOC phenomenon (MOOCS: still not dead). "Tutor Beta breaks OpenStax’s textbooks into smaller chunks, testing students with short answer and multiple-choice questions. The platform also feeds information about how students are learning to instructors." They want to integrate with a "major learning management system" as well. And there's another commercial MOOC trend they have been missing out on: revenue. "Tutor Beta also represents a new revenue stream -- $10 per student per course." Because in the brave new world of open educational resources, nothing says "free" like a subscription charge.
This is an interesting article. It shows you can't just set goals for diversity, you have to work at it, and constantly reinforce the reasons why you're doing it. It also shows that the steps taken to ensure diversity also help ensure a proper hiring process overall - for example, casting a wide net instead of relying mainly on referrals (because people tend to refer people who are like themselves - something I've seen firsthand). It also underlines the value of clear evaluation criteria, and not simply a vague concept of 'cultural fit'. The effectiveness of diversity movements will impact the value proposition for collges and universities, and especially elite institutions. Though they advertise 'great content and great professors' what they're really selling is access to the old boys' network, and though the dysfunction is there for all to see, they won't surrender their poisitions (or their entitlement) without a fight.
According to the website, "Hyperledger Fabric is a platform for distributed ledger solutions, underpinned by a modular architecture delivering high degrees of confidentiality, resiliency, flexibility and scalability." Basically, it's a mechanism for implementing biusiness functions supported by blockchain for security and verifiability. "Hyperledger Fabric 1.0 offers a modular architecture allowing components, such as consensus and membership services, to be plug-and-play. It leverages container technology to host smart contracts called 'chaincode' that comprise the application logic of the system." Now of course it's years between the development of a cross-platform specification like this and widespread implementation, but at the same time, development work in learning technology that is beginning now will implement this approach (if not this specification exactly, then something very similar).
This is an IEEE working group that has receively been reinvigorated. It "defines methods for storing and retrieving learning objects for remote laboratories. The standard will also define methods for linking learning objects to design and implement smart learning environments for remote online laboratories." For example, it defines " interfaces for devices connected to user computers over computers networks and the devices themselves."
This presentation describes some of the work on the IEEE Learning Technology Standards Committee (LTSC). There's downloadable audio and video, and also an embed version that's too slow to try to put into the newsletter. The actual presentation starts at 3:50 of the video and the audio is good. The three major messages: AI is coming, data interoperability standards will be required, and we need to rethink school and learning ahead of the disruption. The video shows only the slides (which include a little bit of embedded video).
Just as an aside: at one point Barr described an algorithm that detects emotions based on facial recognition. Though it has been argued that there are universal expressions emotions across cultures (c.f. Ekman's Atlas of Emotions), other research is arguing that this is not the case. Indeed, it has been argues that people experience emoptions differently across cultures. So we need to exercise caution when we design AI to detect emotions, ensuring at a minimum that training data is cross-cultural and representative.
As the teaser says, "Come for the reinforcement learning, stay for the GIFs." Stay for the short video, actually, to watch a human-shaped robot figure navigate complex landscapes. This aerticle summarizes the outcomes of a reserach paper (14 page PDF), It wasn't given instructions on how to do this; it learned how to jump and climb via reinforcement learning. "It's clear that DeepMind is using creative solutions to get around the obstacles it's presented with; much of the time, the movement that provides the most efficient solution isn't exactly natural looking."
This newsletter is sent only at the request of subscribers. If you would like to unsubscribe, Click here.
Know a friend who might enjoy this newsletter? Feel free to forward OLDaily to your colleagues. If you received this issue from a friend and would like a free subscription of your own, you can join our mailing list. Click here to subscribe.
Copyright 2017 Stephen Downes Contact: email@example.comThis work is licensed under a Creative Commons License.