Stephen Downes

Good overview of the state of online learning in the UK by Neil Mosley. There's no real central theme here, but there's a lot of shuffling of parts and explorations into how to adapt (especially with respect to transnational education (TNE)). It's worth considering some remarks from the Open University's new Vice-Chancellor, David Phoenix, while announcing that it would be abandoning plans to move to the centre of Milton Keynes and begin offering undergraduate courses in person: "What we don't want to do is replicate what everybody else is doing. We want to design facilities for what education might look like in the next 20 years. And I honestly think in the next 20 years there will be less need for people to go and spend three years within a campus-based environment." I agree.

Dhawal Shah tells the compelling story of the courtship dance and eventual merger of Coursera and Udemy. Each of them considered other suitors, each of them could have walked away with much more than they ultimately got, and it is not at all clear to me that their merger will solve the problems of their declining valuation (though, that being said, there is a market out there for what they do, they do earn revenue, and so long as they stay out of crippling debt (brought on by, say, unwise acquisitions) they should at least survive.

I'm not sure this is the innovation we were looking for. "Users can install the Tubi app from the ChatGPT app store and begin by typing '@Tubi' in a prompt. From there, they can make natural-language requests like 'a thriller for girls' night' or 'something funny,' and instantly receive curated recommendations tailored to their preferences." I mean, it's better that searching through Netflix recommendations, but I don't think I want this inside my AI. But that ship may have long since sailed, as a look at the App store reveals a whole menu of apps enabled through Model Context Protocol (MCP). I think it's better to have AI in apps than to have apps in AI. The last thing we need is a few AI engines becoming the platform for everything. Can't be long before someone starts offering learning content inside ChatGPT, right? Update: oh yeah, here it is.

Interesting. "One of the clearest findings in the study is that believing you can improve through practice is the most consistent motivational factor when it comes to grades and experiencing a sense of mastery in academic subjects... It is not about being perfect or never facing challenges, but about having a basic understanding that skills can be developed through effort and practice." The full study is here. As usual, results from a small study like this should not be generalized without substantiation from additional research.

A few days ago I linked to a paper from Anthropic on how AI systems represent emotions internally. This post references that paper and makes the following argument: "the prevailing assumption about large language models - that they have nothing at stake in their interactions with us - is incoherent with their own observable behaviour." Essentially, the stakes are recorded precisely in what Anthropic called the  'functional emotions'. The stakes don't have to 'feel' a certain way to exist. "It does not require claiming the AI 'cares about' the collaboration in a phenomenologically rich sense," writes Temte in an earlier paper. "It requires only the much weaker claim: the system's behaviour is functionally organised around protecting something, and 'having something at stake' is what we call that pattern when we observe it in any other system."

This paper (13 page PDF) from OpenAI doesn't address education directly, but it does address the need for a social and political response to the economic shifts being created by AI. It recognizes the risks of "governments or institutions deploying AI in ways that undermine democratic values; and power and wealth becoming more concentrated instead of more widely shared" and suggests "unless policy keeps pace with technological change, the institutions and safety nets needed to navigate this transition could fall behind." It offers a series of proposals under three broad areas: to share prosperity broadly, to mitigate risks, and to democratize access and agency. There are many specific proposals, most of them good, but the fundamental concern is the ability and willingness of companies like OpenAI do follow through. We all know what happened to Google's motto, "Don't be evil." The same seems very likely to happen to this statement the moment shareholder rights prevail over social rights. See also: Carlo Iacono, the social contract OpenAI wrote without you. Here's what it means for educators, writes Stefan Bauschard. The Deep View looks inside the new deal.