Stephen Downes

It would be hard to disagree with anything in this Charter, released over the weekend by Daniel Supernault, the creator of Pixelfed, though I guess Meta has issues with it, as it has started blocking links to its decentralized competitor (as 404 reports, "Pixelfed is an open-source, community funded and decentralized image sharing platform that runs on Activity Pub, which is the same technology that supports Mastodon and other federated services"). Tyler's comment on 404 summarizes my view: "It's probably within their rights but it always feels so childish when companies throttle links to their competitors. Clearly they don't have enough faith in their product to compete in the 'free market' so they have to cheat." Meanwhile, organizations can commit to uphold the pledge (all pledges are reviewed before being publicly listed). As for me, I'll just indicate my general support here (and note that I am very much following these principles as I design CList, fwiw).

Discussions of ethics of technology in learning often focus on the three schools of ethics - virtual theory, consequentialism, and deontology. They rarely mention contractualism, which tells me simply that their foundations in actual ethical theory are slight. This essay, a well-written and accessible review of T.M. Scanlon's collection of essays, Morality and Responsibility (itself behind a paywall, sadly), helps address that deficiency. Scanlon's work reshapes some core assumptions about ethics: the idea that what counts as 'ethical' is contextual and based in relationships, and the idea that attributions of responsibility and consequences do not require a foundational theory of free will. "An action or policy is wrong," says Scanlon, "if any principle that permitted it could be reasonably rejected by someone affected adversely by that principle."

I like the way Miguel Guhlin is recasting Epictetus's Enchiridion, a core work in Stoic philosophy. He sets up the series with this podcast episode and then settles in with the actual chapters (three of six so far: one, two, three). "It opens with the Stoic premise stated plainly: 'some things are up to us, and some things aren't. What's up to us is our own perspective, our goals, and our choices. What's out of our hands is our health, our wealth, and our reputation.'" Or, put another way: "Look, the person who can give you what you want, or take away what you're afraid of losing - that person is your master. If you want to be completely free, stop wanting things from other people and stop fearing what they can do. Otherwise, you're just a slave in a nice suit." Now, of course, there are things we can do that influence our health, reputation and wealth - we don't live in the year 135 any more. But the advice - that we should focus on what is within our control - is still good.

I love how this article presumes it will be a company that creates an alternative to university credentials, and not (say) a government, a cooperative or a community. Presumably that's because the idea here is centered around a product: "This new product would be a registry, a kind of credit bureau, land registry, and LinkedIn all at once. The registry confirms you learned something and registers it. The world can check the registry when it needs to know what the student can do. An examination will surely be involved but the registry, not the examination, is the asset." But why would it be a (centralized) registry? One of the strengths of the university system is that it is decentralized - you can't simply replace it in one go, and it doesn't fail with the collapse of a single institution. 

Though sceptical of AI, Maha Bali describes in this post her efforts to "encourage them (students) to use their own judgment to see where AI can support them in doing the work, without replacing them or limiting their creativity, their voice, their identity." And if they do use AI, she says, "I just ask for transparency." I think this is a good approach, and as this article documents, she was rewarded in turn. "Students are getting better at using AI, not in the sense of getting better at prompting it to do more work, but in the sense that they are becoming better judges of what appropriate use looks like and what inappropriate use looks like. And able to self-control and take agency over their use of AI when I gave them freedom beside critical AI literacy." 

"Teaching AI Literacy and Preserving Assessment Integrity are not the same problem," writes Mike Kentz. This should be obvious but apparently the issues are being conflated in meetings convened to address 'the AI problem'. "The key insight: assessment integrity can be preserved or rebuilt without involving AI literacy at all. AI literacy can be built without navigating around AI cheating. These are separate tracks aiming at separate outcomes." For those who are wondering: breaking complex problems down into simpler parts is known as the Cartesian method, and was introduced in his Discourse on Method (1637).