Stephen Downes

I have virtually no opinions in common with Paul Wells, but I have addressed the need for universities to become more relevant to communities, and this post seems to address that. It reads though more like a thinly veiled warning that universities are going to have to align politically. For example, "The easiest way to 'force' universities 'to do better' is to 'review the eligibility requirements for the receipt of federal research funds to ensure strong university governance'" which would more directly address, say, protest encampments. To me, this sort of argument underlines once again the need for universities to become essential in the lives of the people in the community, because only the community can support the university when its funding is being pressured.

I found this game more annoying than fun, and I have no idea why they would need to use the Unreal engine to design it. It reminded me of the old-style Flash games. Anyhow, as Nele Van de Mosselaer says, "Take, for example, a door in a videogame. There is nothing fictional about such a door: it is a simulated door that actually exists on your computer. You can see it, as it is there on your screen. You can open it, close it, maybe even lock and unlock it. You don't need to imagine anything, you just interact with it." Which does raise some interesting questions about representation. So, enjoy?

Carys Craig (29 page PDF) argues (and I agree) that "Copyright law should neither incentivize and reward the use of generative AI nor obstruct its training and development." However, "it seems clear that copyright law (or a contorted version thereof) is increasingly being invoked as a regulatory response to the harms of AI." This paper is an extended treatment of the argument. In particular, "The so called "3Cs" of 'Consent, Credit and Compensation' are getting a lot of air time these days." Craig argues, "the pursuit of the 3Cs is intended to push back at power, employing the blunt tool of copyright control, but reaching beyond what copyright actually requires by narrowing the scope of what fair use permits (in the name of greater fairness)."  We have to remember that these limits on copyright are "limits that have traditionally restrained corporate power to protect the public interest... copyright is entering the fray as a false friend."

So the way this study (59 page PDF) worked is that a teacher taught the students, then they had some AI-assisted practice, then they took a test on the same material with no AI support. They performed worse when they used the AI than when they didn't. I found this a pretty narrow study, and I'm not really sure about the AI the researchers used (it was a ChatGPT 4 'base' model, which they report frequently made math errors). I wouldn't think simply using the AI for a 'practice session' would be particularly interesting or engaging, but maybe that's just my perception.

This article summarizes a report Teen and Young Adult Perspectives on Generative AI and states the main points as follows (quoted):

• Half of teens surveyed have used generative AI, but few (4%) use it frequently
• Most common uses are for getting information (53%) and brainstorming (51%)
• AI can help with homework and creative assignments, not just cheating
• Students bring their questions to generative AI, for better or worse
• Many students are worried about AI used for bullying and disinformation 

None of these should be surprising, except perhaps that so few students report using AI daily.

Publication of the study is still pending, but according to this article the use of an AI tutor in a physics course greatly increased engagement and with it the amount of learning. "'We went into the study extremely curious about whether our AI tutor could be as effective as in-person instructors,' Kestin, who also serves as associate director of science education, said. 'And I certainly didn't expect students to find the AI-powered lesson more engaging.' But that's exactly what happened."