OLDaily

Welcome to Online Learning Daily, your best source for news and commentary about learning technology, new media, and related topics.

The AI-Powered Nonprofits Reimagining Education
Kevin Barenblat, Brooke James, SSIR, 2024/08/26

---

According to this article, "AI is being used in exciting ways to bridge educational divides, and AI-powered nonprofits are creating a roadmap for what the future of education may hold." In particular, the article offers three case studies: Rocket Learning, which uses AI to create localized content in various Indian languages; Khanmigo, a Khan Academy teaching assistant; and Learning Equality, which uses AI for curricular alignment and lesson development. Now these are probably not the best three examples - after all, they appear in SSIR, which has an investor and capitalist focus. But they are indicators of what can be done, and most likely will be done.

Web: [Direct Link] [This Post][Share]

My AI Breakthrough
Miguel Guhlin, Another Think Coming, 2024/08/26

---

How does one reconcile the argument that AI is disempowering with this: "something that took me WEEKS of hard work, and in some cases I found impossible, was made easy. Like, instead of weeks, it takes 10 minutes." People who are actually using AI are reporting this sort of result over and over. Not for all work, and not in all cases. But enough that to me there is a compelling argument to be made here.

Web: [Direct Link] [This Post][Share]

Big mistake — Schools are swapping out Shakespeare, Chaucer and Dickens for Kendrick Lamar and Taylor Swift
Liza Libes, The Hechinger Report, 2024/08/26

---

This is why you shouldn't depend on consultants to teach writing classes. No, not because they're teaching students about Taylor Swift and Kendrick Lamar - after all, if Bob Dylan can win a Nobel Prize, popular culture can be literature. No, it's this: "I teach writing to high schoolers through my college consulting firm," writes Liza Libes. "When preparing students to write their college essays, I always lead with literature — Hemingway, Dostoyevsky, Fitzgerald, for instance — for I believe that developing strong communication skills lies in understanding the writing styles and messages of these great thinkers and writers." How useless would this be?

Web: [Direct Link] [This Post][Share]

Harvard and MIT's $800 Million Mistake: The Triple Failure of 2U, edX, and Axim Collaborative
Dhawal Shah, The Report by Class Central, 2024/08/26

---

This is a story of unmitigated failure. In 2021, "the unprofitable 2U bought edX, an unprofitable non-profit, for a staggering $800 million." The resulting entity is now bankrupt, an outcome that even in foresight seemed to be inevitable. As Dhawal Shah writes, "2U's business model required (online) degrees to be as expensive as on-campus degrees." Meanwhile, " Axim Collaborative is the non-profit organization that retained the ~$800 million from the 2U acquisition. It's the only cash-rich entity in this story, yet three years later, it has little to show for its wealth."

Web: [Direct Link] [This Post][Share]

A symbol for the fediverse ⁂
⁂ fediverse symbol, 2024/08/26

---

"We propose the symbol ⁂ to represent the fediverse," write the authors. The advantage here is that the image is a character in standard character sets, and so we don't need to insert an image every time we want to use it. Am I excited by it? No. Is it practical? Yes. Will I use it? Probably. Via Ben Werdmuller.

Web: [Direct Link] [This Post][Share]

The thorny problem of authorship in a world of AI
Doug Belshaw, Thought Shrapnel, 2024/08/26

---

"As Tunney points out," writes Doug Belshaw, "the world of Open Source is a gift economy. But if we're gifting things to something ingesting everything indiscriminately and then regurgitating in a way that erases authorship, is that problematic?" Tunney writes, "if these AIs like Claude are learning from my code, then what I want is for Claude to know and remember that I helped it. This is actually required by the ISC license." Looking at my own learning, I would find it impossible to credit everyone I learned from in order to create, say, this post. Sure, where I'm directly quoting someone, I can credit them. That's a trivial problem AI could easily solve when it directly quotes someone. But if AI learns the phrase 'points out' from, say, 500 different examples, does it make sense to credit each of them?

Web: [Direct Link] [This Post][Share]

We had an AI attempt to make a data-driven story like we do at The Pudding
Russell Samora, Michelle Pera-McGhee, The Pudding, 2024/08/26

---

The answer to the question is "Yes, but not as well." The AI did quite well at the simpler and more focused tasks, but more poorly at the more creative tasks involving longer narratives or chains of thought. "Do we feel replaceable? In short, not right now. When we look at what Claude produced, we don't think it looks anything like a real story we'd make at The Pudding. But… would a casual reader notice the difference? We hope so, but that might be overly optimistic.

Web: [Direct Link] [This Post][Share]

The Soul of Open is In Danger
Heather M. Ross, McToonish, 2024/08/26

---

Responding to the description of an upcoming talk by David Wiley, Heather Ross argues strenuously against the blending of artificial intelligence and open educational resources. "GenAI may be fun to play with and make some tasks easier," she writes, "but the cost to the values of open, the planet, marginalized groups, and humanity as a whole are far too great." Now I disagree with her argument, and not merely because I made the same argument Wiley is making several years ago. But even more, as someone who has been building and developing this field for 25 years, I strenuously object to language like "We can't stop you, but get off our field." Image: ChatGPT/DALL-E.

Web: [Direct Link] [This Post][Share]

Andy Jassy on using generative AI in software development at Amazon
Ben Werdmuller, Werd I/O, 2024/08/26

---

Does AI work? "The average time to upgrade an application to Java 17 plummeted from what's typically 50 developer-days to just a few hours. We estimate this has saved us the equivalent of 4,500 developer-years of work (yes, that number is crazy but, real)." My own experience tells me that this is certainly plausible. Moreover, the resulting product is better, with "enhanced security and reduced infrastructure costs." Obviously this is an outlier statistic. But it's a very strong indicator of the direction in which this is headed.

Web: [Direct Link] [This Post][Share]

Generative AI chatbots in higher education: a review of an emerging research area - Higher Education
Cormac McGrath, Alexandra Farazouli, Teresa Cerratto-Pargman, Higher Education, 2024/08/26

---

Good survey paper discussing recent research on AI chatbots in higher education. While there are many well-documented concerns, the studies find high levels of use, resulting in a "love-hate" relationship between students and AI. The authors write, correctly, that "it is important to observe the social impact of using GAI chatbots in HE as a social arena, not just as a site for student learning and teaching." And "the scientific literature needs to resist contributing to the dystopian and utopian hyperbole when framing and considering the transferability of research findings."

Web: [Direct Link] [This Post][Share]

How to win at CORS
Jake Archibald, 2024/08/26

---

I recently wrote a simple data store web application in Python (in the process learning a lot of Python) connected to a separate Javascript only client on a different server. In the process I learned a lot more than I ever wanted to about cross origin resource sharing (CORS). This is a set of standards that allow services from different websites to interoperate but without compromising security. This is a great article from a couple of years ago on CORS, just shared today by Simon Willison. If you're developing web technology, you need to know about CORS.

Web: [Direct Link] [This Post][Share]

We publish six to eight or so short posts every weekday linking to the best, most interesting and most important pieces of content in the field. Read more about what we cover. We also list papers and articles by Stephen Downes and his presentations from around the world.

There are many ways to read OLDaily; pick whatever works best for you:

• Read in your web browser on this web page, updated weekdays
• Receive a daily or weekly email newsletter by subscribing here
• Subscribe to the RSS feed using your favourite feed reader
• Follow @OLDaily@mastodon.social on Mastodon or elsewhere in the fediverse
• Follow @OLDaily on Twitter
• Integrate the open API using the JSON feed