OLDaily

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

ExplanAItions 2025: The Evolution of AI in Research
Wiley, 2026/02/19

---

I came across this (53 page PDF) while doing some desk research and though it's from last year (and nominally behind a spamwall) I thought I'd pass it along. The real value of this report is the 44 use cases for AI in research that it lists and organizes according to how likely and how useful they are (with a number falling into the 'humans preferred' category). There's also what they call the "Wiley AI Framework" describing what people should act on, watch, or envision. It's also interesting to look back on their perspective from about twelve months ago: "This year, GenAI enters the Trough of Disillusionment as organizations gain understanding of its potential and limits. AI leaders continue to face challenges when it comes to proving GenAI's value to the business." The survey does reflect that for 2025, but for 2026 I think we're already seeing significant advances.

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

Diamond Open Access Needs Institutions, Not Heroes
Curt Rice, The Scholarly Kitchen, 2026/02/18

---

This is a good argument. Diamond open access refers to academic texts that are published, distributed and preserved with no fees to either reader or author. Curt Rice points out that it used to exist in the pre-internet era: "Manuscripts circulated through departmental working papers series and informal scholarly networks." I remember those days. Rice writes, "What made this ecosystem possible was not heroism, but tractability: limited scale, manageable volume, and informal governance. And that is precisely what has changed." A lot of the commercial publishing infrastructure developed as a way to try to account for this scale, but it has failed as a scholarly activity. "Paying reviewers or editors reframes scholarly contribution as a transactional service rather than a professional responsibility embedded in institutional roles... More importantly, payment does not solve the problem of alignment. What many academics seek is... assurance that their professional contributions are recognized, supported, and valued within the institutions that depend on them." This makes sustainability for diamond open access publication an institutional responsibility, and underlines the need to built structures that supports it.

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

Agentify Your App with GitHub Copilot’s Agentic Coding SDK
Shittu Olumide, Machine Learning Mastery, 2026/02/18

---

OK, forget about all the tech in this post (and there's a lot of tech in this post). Take a look at the diagram and reflect on what it says about thinking generally. In the past we've had things like critical thinking and computational thinking. These were useful concepts. We might now want to coin a new discipline 'agentic thinking' to follow this sort of model, summarized as "autonomous execution, multi-step problem solving, persistent context, tool use," and deploying skills such as 'task planning', 'tool orchestration', 'multi-turn conversation' and 'evaluation'. None of this is new, per se, but organizing our approach to creativity, problem-solving and decision-making is, I think.

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

On Student Success
Glenda Morgan, On Student Success, 2026/02/18

---

I'm mentioning this here so people know that this relatively new newsletter exists, and also to urge Glenda Morgan to add an RSS feed so I fan follow it in my RSS reader. It's a sibling publication to Phil Hill's On Ed Tech, which mentioned it today. Hill's posts are often for subscribers only, which is why I cite it less often than I might, but I always read the stubs on his RSS feed, just to keep track.

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

The Hidden Dangers of Meta’s Partnership Offer to Schools
Faith Boninger, Progressive.org, 2026/02/18

---

"In April 2025, Meta started recruiting U.S. middle and high schools to participate in Instagram's new School Partnership Program, inviting schools to partner with Instagram to help combat online bullying," reports Faith Boninger. This is an arrangement more likely to benefit Meta than the schools, she writes, by enlisting schools in their (limited) content moderation efforts. And "the idea that Meta will review content reported by partner-school accounts sooner than other reported violations also implies a veiled threat: that schools that do not partner with the company will find themselves waiting longer for review." Schools should disassociate themselves from Meta, she says. "Yes, many kids will still use Instagram. But at least their school won't be leading them there." Via Larry Cuban. Related: Meta plans to add face recognition to its smart glasses.

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

The Civic Stakes of Organizational Disagreement
Peter Levine, Dayna L. Cunningham, SSIR, 2026/02/18

---

I mostly agree with what's in this article, not simply because disagreements within organizations are essential for democratic processes, but because (as I've often said) diversity is necessary for any organization to learn and adapt. There's a nice way of putting it buried in the centre of this article: "If neutrality is impossible, pluralism is an essential ideal." It doesn't matter whether we're talking about a person, a company, a university or a society: we have to make decisions. We cannot remain neutral (whatever that even means). And in such a case, the best and only reasonable approach is to consider various possibilities and negotiate our way to a resolution. These are never final; we have to do it each time. And as the article notes, the trick is to do it without rending ourselves in the process. "Hannah Arendt argued that a good life requires consequential debate among equals who are meaningfully different; Jürgen Habermas insisted that collective legitimacy depends on free, inclusive, and reasoned deliberation." This isn't just the responsibility of leaders. Everybody has a role to play. That's why, in any company, university or society, everybody is important.

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

A Guide to Which AI to Use in the Agentic Era
Ethan Mollick, 2026/02/18

---

The developments in AI technology continue apace. Today's new concept to learn is the 'harness'. Ethan Mollick writes, "A harness is a system that lets the AI use tools, take actions, and complete multi-step tasks on its own. Apps come with a harness." He continues, "Until recently, you didn't have to know this. The model was the product, the app was the website, and the harness was minimal. You typed, it responded, you typed again. Now the same model can behave very differently depending on what harness it's operating in." This makes a big difference in how an AI works for you, so you need to check. If you're using ChatGPT 5.2, for example, are you using 'auto', 'thinking' or 'instant'? It depends on what you're using it for.

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

Current
Terry Godier, 2026/02/18

---

What I love about the arrival of AI-supported coding is that now people can create their own flavours of old standards. Take RSS readers, for example. Take Current, for example. "Each article has a velocity, a measure of how quickly it ages. Breaking news burns bright for three hours. A daily article stays relevant for eighteen. An essay lingers for three days. An evergreen tutorial might sit in your river for a week. As items age, they dim. Eventually they're gone, carried downstream. You don't mark them as read. You don't file them. They simply pass, the way water passes under a bridge." My needs are different; I wanted to surface unread voices, so I wrote a reader that displays those who post least frequently first. I also wanted itto run locally, but sync across instances. So that's what I wrote. Each person gets their own version, their own flavour, of what they want.

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

Why the Difference Between AI Literacy and AI Fluency Matters
David Ross, Getting Smart, 2026/02/17

---

This is a useful article because it links to a number of recent AI literacy frameworks. It also makes the case that our mastery of AI will need to go beyond literacy. "I suggest we adopt a tried-and-true educational model that governs our thinking around curriculum frameworks: A scope and sequence. The final outcome of that scope and sequence should be AI fluency." This could be true - but I would caution that what counts as AI literacy, much less fluency, is very much a moving target. As a case in point, for example: do we or don't we need prompt engineering? New, related, and not mentioned in the article: Acadia University's free Introduction to AI Literacy course (and CBC coverage). Also: the U.S. Department of Labor's new Framework for Artificial Intelligence Literacy (15 page PDF).

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

Our Emerging Planetary Nervous System
Rimma Boshernitsan, NOEMA, 2026/02/17

---

This is a longish article with some good examples showing a future state (and likely applications) of what the author calls our planetary nervous system. What is meant by that is the interconnected network of sensors and indicators that respond to what's happening in the natural world, with inputs ranging from waterflows to migration patterns to the spread of wildfires. "This is machine intelligence at its most vital," writes Rimma Boshernitsan, "not replacing judgment, but extending our senses." The objective is "to integrate so coherently with the biosphere that the whole can self-regulate rather than just react." And what we want, I would say, is for this integration to be available to everyone, the way the Global Biodiversity Information Facility (GBIF) "weaves millions of records from field notes, museum collections, citizen observations and satellite traces into a living archive" creating "a global network and open-access infrastructure funded by governments worldwide." If we don't require that this data be open access, someone will attempt to privatize it.

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

Agentic Email
Martin Fowler, martinfowler.com, 2026/02/17

---

Martin Fowler reports, "I've heard a number of reports recently about people setting up LLM agents to work on their email and other communications. The LLM has access to the user's email account, reads all the emails, decides which emails to ignore, drafts some emails for the user to approve, and replies to some emails autonomously." As enthusiastic as I am about AI, I agree with him that it's far too early to trust agentic email with real access to my email, and not only because of the security risks. I mean, I can't even trust my anti-spam services to keep out all the spam and only the spam. I'm not ready to let it make statements on my behalf. And oh yeah, the security risk.

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

Top Priorities for Global Heads of Learning and Talent
iVentiv, 2026/02/17

---

This report (26 page PDF) is based on a survey of 468 heads of learning and talent from 394 companies in Europe, the United 
States, the UK and the Middle East. The results are not surprising. The top priority continues to be leadership and executive development, as ever, and close on its heels is artificial intelligence. Also, "In last year's data, the phrase 'skills-based organisations' came up more than twice as often as in 2024.... That persistence reflects the 'skills based' approach becoming part of the 
mainstream as we head into 2026." This priority, and also the emphasis on 'learning culture', reflects the need to adapt to a rapidly changing skills landscape; for this reason as well it is difficult to link learning programs directly to return on investment (ROI). The conditions before and after learning are often completely different, and it's often a case of 'adapt or get left behind' for individuals and companies. The report is behind a spamwall, and you can give them your contact information if you're feeling nice, but this direct link should work as well.

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 Bluesky
• Read OLDaily as a LinkedIn Newsletter
• Integrate the open API using the JSON feed