Stephen Downes

I tried the Lightcraft application on ArtCraft - it's an attempt to recreate Lightroom. It was unable to import my RAW .NEF files from my camera, so it allowed me to edit the included .JPG instead. But that's not really a substitute. I notice as well that performance is similar to other non-Adobe editing apps, such as Affinity. So there's a way to go. Still. It's hard to disagree with the headline that "software is over". Tom's Hardware has a good overview of the project. "Excited testers say this could be a legitimate Adobe alternative, but others claim it may just be more vibecoded AI slop. It remains to be seen how many will collaborate to improve upon this project and fill in the missing features that it plans to replicate." It might take more than a month. Maybe even a year.

I'll just quote this, with approval: "universities inevitably reflect social values and institutional histories, which is precisely why claims of neutral meritocracy deserve critical scrutiny. The real threat to academic excellence is not Equity, Diversity and Inclusion (EDI). It is the persistence of the meritocracy myth: the comforting but inaccurate belief that universities have always been neutral arbiters of talent. EDI challenges that myth by insisting that excellence is too important to be constrained by historical patterns of exclusion. Rather than lowering standards, EDI asks universities to broaden their understanding of where excellence can be found and how it can flourish." If I had any additional remarks, it is to recommend that our understanding of EDI should be sufficiently broad as to include socio-economic status (SES) as one of its criteria. 

The work I've done on CList is about creating social interacting learning networks based on personal data spaces and environments, so there's a lot of overlap with ActivityPods, which "Brings together ActivityPub and Solid Pods and empowers developers to create truly decentralized applications." As described here, "all your data are stored in your Pod. You can choose a Pod provider you trust, or host it yourself....  you can share your data with anyone on the Internet, thanks to the ActivityPub protocol." And "since all data are stored using semantic web standards, they can be accessed and understood by many apps." Really interesting. I think one of the big difference between CList and this sort of project is that while you can store or host data yourself, you don't have to. Anyhow, here is the ActivityPod documentation.

Google's Gems are "customized versions of Gemini that help you tackle repetitive tasks or get deep expertise in new areas." They were used by teachers in Google Classroom. But now, as this article reports, Google is retiring Gems and replacing them with skills. But skills won't be available in Classroom. So "what happens to Gems you assigned to students, the key dates, and what to move them to." Similarly, "OpenAI retires custom GPTs on December 11, 2026" - there's an article on what teachers were building with them." Once again, we are learning the danger of relying on a proprietary platform. Now this article promotes something called Flint as an alternative; I can't vouch for them, but at least they posted this useful article.

This I think is true: "digital sovereignty is not achieved by owning more of the technology we depend on but rather by having meaningful agency over it and the capacity to exercise that agency over time." This isn't something that should depend on volunteers, writes Anna Tumadóttir, but is something that needs to be sustained. But how? The warning here is that "Open infrastructure projects can become dependent on philanthropic funding or on the priorities of a particular government or administration." What's the best way to sustain infrastructure generally? We can't depend on the private sector for this, because the private sector prioritizes extracting revenue, not supporting the infrastructure. How often we have seen this! That's why I worry about the direction Creative Commons is taking., as it drifts closer and closer to the idea of a private sector controlled commons. Most key infrastructure is supported through public enterprise - in Canada, that means through government agencies, institutions, or Crown corporations. That way, if they're being mismanaged, we can replace the managers.

In his talk at FediForum yesterday Anil Dash outlined a process whereby companies spend a lot of money to take over a niche where open content prevails, then slowly squeeze out all the openness, then monetize it as a closed proprietary sector. A classic case of this is Spotify, which bought up a bunch of podcasting companies, and has been squeezing ever since. This item is a case in point. Spotify has begun scanning podcast episodes for any spoken work content that might come from a commercial source. If it finds any, the episode is removed. There's no fair use or any legitimate way to even quote from a work. "This is just another example of the ruination of the legal concept of fair use. Fair use is not a copout or a loophole, it's a fundamental understanding that copyright law has limits, and quoting a work to criticize it can't be stifled by authors demanding that their copyright entitles them to prevent critics from quoting their work." See also Anil Dash, We are going to kill "unalive".