Sign in

Stephanie Boragina

@sboragina.bsky.social
729 followers 1.2K following 87 posts

🇨🇦 PhD student and instructional designer of online courses Studying math education at the postsecondary level. Currently examining the growth of student mathematical understanding when taking asynchronous online math classes. #MathEd #OnlineEd

PostsRepliesMedia
Reposted by Stephanie Boragina
Jeff Greene @jeffgreene.bsky.social · 16/06/2026
Good discussion of the roles of habits in learning, how to intervene on those habits to promote better learning strategy use, and ideas for future research on the role of habits in self-regulated learning. doi.org/10.1007/s106...
doi.org
Understanding Learning Strategy Use Through the Lens of Habit - Educational Psychology Review
Students frequently rely on ineffective learning strategies instead of those that promote long-term retention. This is not simply a matter of lacking metacognitive knowledge. Research on…
043
Reposted by Stephanie Boragina
Sonja Drimmer @sonjadrimmer.bsky.social · 14/06/2026
One of the things I research is the cooperation of cultural institutions like museums in pitching tech products--specifically AI--to the public. This is advertising for the capacity of AI to educate and thus to reshape education itself as a project better left to private tech companies. 1/n
12442162
Stephanie Boragina @sboragina.bsky.social · 20/06/2026
Have been thinking about CS education and AI use recently, and came across this article in Nature about AI and deskilling. www.nature.com/articles/d41... Related to CS education, in particular, it highlighted this recent study (pre-print available).
No lesson learnt
To investigate whether skills are being lost in the field of computer science, researchers at the Al firm Anthropic in San Francisco, California, designed a randomized controlled trial in which 52 software engineers were asked to perform a basic coding task³. During the exercise, all 52 participants could search the web and access instructions on how to do the task. Half of the participants were prompted to use an Al assistant as well.

Afterwards, all of the software engineers were asked to complete a quiz about what they had learnt
from the task. The participants who had used an Al assistant did significantly worse on the quiz than those who hadn't: the average (continues on next image)score was 50% in the Al group versus 67% in the non-Al group. The Al-assisted participants did particularly poorly on questions that required them to diagnose errors in the code, which suggests that they had failed to learn the concepts behind the code that they had just produced. The study was posted on the preprint server arXiv ahead of peer review.
The findings are of concern, especially for students and young professionals in the field, says Crowston, who is researching how the use of generative Al tools is changing the way that software developers learn and retain coding skills. "Now you have this very odd disconnect between performance and learning," he says. “People can perform at a pretty high level, because they're basically borrowing skills from the Al, but they are not developing those skills themselves."
141
Stephanie Boragina @sboragina.bsky.social · 02/05/2026
TEP- AIED is at the intersection of

Transparency
- System specification
- Prompt design
- Reproducibility

Ethics
- Data governance
- Risk mitigation
- Equity and access

Pedagogy
- Learning objectives
- Al instructional role
- Learner orientation
010
Reposted by Stephanie Boragina
Derek Bruff @derekbruff.bsky.social · 01/05/2026
Key line: "No structured tutorial or prompt-engineering guidance was provided; the intention was to capture AI use as it naturally occurs, with participants free to interact as they typically would." Contrast with the well-prompted tutorbot in the Kestin study: www.nature.com/articles/s41...
342
Reposted by Stephanie Boragina
Petter Törnberg @pettertornberg.com · 01/05/2026
LLMs have been widely reported as left-wing biased. The finding has shaped policy and debate — with Trump banning "Woke AI". Our new paper challenges this story. It's not that the models are biased. It's that they think the auditor is. 🧵 w/ michelleschimmel.bsky.social‬arxiv.org/pdf/2604.27633
415750
Reposted by Stephanie Boragina
Dr. Mag 🌿👾 @nobugsnous.bsky.social · 11/04/2026
I hope this helps people move away from process tracking. it feels worth pointing out that anyone can manually type out LLM outputs.
195
Reposted by Stephanie Boragina
David Hopkins @hopkinsdavid.bsky.social · 09/03/2026
“So the hollowing out of PS staff is a massively false economy. It transfers work from relatively low-paid PS staff to relatively highly paid academics, who are not renowned for their administrative efficiency.“ www.linkedin.com/pulse/how-un...
linkedin.com
How universities have been reorganised into incompetence
The people who administer the work of universities are known as professional services staff, or PS for short. As if their work was some sort of additional postscript to what universities do.
041
Reposted by Stephanie Boragina
Sandra Abegglen @sandra-abegglen.bsky.social · 13/03/2026
Great article exploring how GenAI reshapes stance & voice in student writing: www.degruyterbrill.com/document/doi... Findings indicate that multilingual student writing is drifting toward the stylistic defaults of GenAI even before the tool is applied, & polishing then reinforces this convergence.
degruyterbrill.com
Beyond Polishing: The Compounding Dynamic of GenAI in Academic Writing
Generative AI can reshape stance and voice in multilingual student writing, with effects that appear to be shifting over time. Using a random sample of English L2 master’s-level assignments at a UK un...
021
Reposted by Stephanie Boragina
Michael Barger @mmbarger.bsky.social · 13/03/2026
New short report out from my team (Shannon Clark, Darren Henry, Jordan Rineer)! We asked elementary teachers what kinds of responses they tend to give their students in math 1/7 link.springer.com/article/10.1...
153
Reposted by Stephanie Boragina
Jeff Greene @jeffgreene.bsky.social · 12/03/2026
We are rapidly approaching the "can't swing a tetherball without hitting one" number of meta-analyses of #GenAI effects on learning. Perhaps we should do better primary research on that topic before trying, yet again, to meta-analyze a bunch of flawed studies? onlinelibrary.wiley.com/doi/10.1111/...
onlinelibrary.wiley.com
ChatGPT in Education: An Effect in Search of a Cause
Background As researchers rush to investigate the potential of AI tools like ChatGPT to enhance learning, well-documented pitfalls threaten the validity of this emerging research. Issues of media co.....
1111
Stephanie Boragina @sboragina.bsky.social · 12/03/2026
Thinking a bit about these "learning pathways."
The pedagogical solution: From cognitive atrophy to augmentation
The evidence for detrimental offloading is compelling, but it is not deterministic. The technology itself does not seal the outcome. As Kalyuga and Plass (2025) argue in their update to Cognitive Load Theory, the goal of the student matters, as does their motivation. Much of the research to date has yet to grapple with this added complexity. As Weidlich et al. (2025) argue in their methodological critique, much of the research on Al is an "effect in search of a cause," because it conflates the medium (the Al tool) with the instructional method (the pedagogy). They argue that pedagogy is the causal factor, and the research provides clear, evidence-informed pathways for a pedagogy that mitigates the risks posed by Al outsourcing. Not only is goal setting (a component of self-regulated learning) critical here, but other aspects of students' capability to manage their own learning and thinking also need to be supported through their learning experiences.
That said, the pedagogical design of the tools used in instruction can influence whether students engage in positive or negative cognitive offloading. Al tools vary in their integration of educational evidence and construction of learning pathways and they increasingly form a mediating layer between curriculum and delivery in the classroom (Loble & Stephens 2024a).
Path 1: Beneficial offloading and load reduction instruction
Al can be used for beneficial offloading, managing extraneous load to free resources for intrinsic learning. This requires an explicit pedagogical framework. Martin et al. (2025) provide such a framework, "Load Reduction Instruction" (LRI), which adapts explicit instruction principles useful for managing learning with Al. Continued from previous image... Drawing on this model, Al can be used to provide scaffolding, structured practice, and feedback, all aimed at managing the cognitive burden on the learner and enabling progressive independence (Martin et al. 2025), in other words, helping students to become better self-regulated learners. As validated by Hong et al. (2025), students explicitly taught this cognitive offload instruction model (offloading low-order writing tasks) showed significantly greater gains in critical thinking.
Path 2: Scaffolding metacognition to counter laziness
The more profound solution is one that directly tackles the core problem of metacognitive laziness (Fan et al. 2024). If the problem is that the convenience of Al encourages learners to abdicate their metacognitive responsibilities, the solution is to design Al interactions that explicitly demand and scaffold those responsibilities. While these design parameters may not be within the direct control of educators, these kinds of prompts can be used in a wide range of scenarios to help students develop these capabilities. It will be helpful when Al tools increasingly have these capabilities built in, but technology is not required for teachers to use these approaches.A compelling cluster of recent studies demonstrates
the success of this approach, building on a longer history of metacognitive prompting research:
Xu et al. (2025) found that integrating metacognitive prompts into an Al environment significantly enhanced learner capability, particularly in terms of self-regulated learning. Singh et al. (2025) found that metacognitive prompts designed to make users pause, reflect, and assess their understanding led to more active engagement, broader topic exploration, and deeper inquiry during Al-based search. Li et al. (2025) used a progressive prompting intervention, where the Al provides gradually fading scaffolds, and found that it significantly improved both learning achievement and critical thinking skills.
These studies show that the passivity induced by Al can be overcome with explicit metacognitive interventions. Technology is not needed to achieve this. This pedagogy resolves a key conflict: Darvishi et al. (2024) found that simply adding SRL prompts failed because the dominance of Al overrode them. The successful interventions (e.g., Xu et al. 2025; Singh et al. 2025) were integrated and non- optional, forcing the metacognitive pause. Many learning activities, with or without technology, can be implemented to achieve this goal.Path 3: Designing Al as a cognitive mirror and verification partner
The most advanced pedagogical and technological design solutions shift the fundamental role of Al from an answer oracle (which invites passive outsourcing) to a tool that provokes intrinsic load.
+ Al as cognitive mirror: Tomisu et al. (2025) propose this framework. The Al is engineered as a teachable novice with a pedagogically useful deficit. It feigns confusion and asks clarifying questions, forcing the human learner into the effortful, generative act of explanation and reflection (the "Protégé Effect"), thus triggering the generation effect (Duplice 2025).
Al as Socratic partner: Monzon and Hays (2025) propose using Al to create desirable difficulties. Instead of bypassing effort, the Al is used as a cognitive partner to generate retrieval-practice questions, case studies, and Socratic dialogues that force the effortful processing required for durable learning.
+ Al as verification partner: Grace (2025) proposes a model of intelligence equilibrium where the human maintains primary cognitive agency and continuously evaluates and corrects the Al output, guided by a verification mindset.
As synthesised by Helal and colleagues (2025), the impact of Al is entirely dependent on the pedagogical design. The default, unstructured answer oracle activates cognitive inhibitors like automation bias and quick-solution dependence. In contrast, intentional, pedagogically sound design activates cognitive mediators like self- regulated learning and metacognitive critique.
020
Reposted by Stephanie Boragina
Dan Chibnall @bookowl.bsky.social · 10/03/2026
Sycophantic AI strikes again. Filed under "not surprising, still frustrating." arxiv.org/abs/2510.01395
Sycophantic AI Decreases Prosocial Intentions and Promotes Dependence
Myra Cheng, Cinoo Lee, Pranav Khadpe, Sunny Yu, Dyllan Han, Dan Jurafsky
Both the general public and academic communities have raised concerns about sycophancy, the phenomenon of artificial intelligence (AI) excessively agreeing with or flattering users. Yet, beyond isolated media reports of severe consequences, like reinforcing delusions, little is known about the extent of sycophancy or how it affects people who use AI. Here we show the pervasiveness and harmful impacts of sycophancy when people seek advice from AI. First, across 11 state-of-the-art AI models, we find that models are highly sycophantic: they affirm users' actions 50% more than humans do, and they do so even in cases where user queries mention manipulation, deception, or other relational harms. Second, in two preregistered experiments (N = 1604), including a live-interaction study where participants discuss a real interpersonal conflict from their life, we find that interaction with sycophantic AI models significantly reduced participants' willingness to take actions to repair interpersonal conflict, while increasing their conviction of being in the right. However, participants rated sycophantic responses as higher quality, trusted the sycophantic AI model more, and were more willing to use it again. This suggests that people are drawn to AI that unquestioningly validate, even as that validation risks eroding their judgment and reducing their inclination toward prosocial behavior. These preferences create perverse incentives both for people to increasingly rely on sycophantic AI models and for AI model training to favor sycophancy. Our findings highlight the necessity of explicitly addressing this incentive structure to mitigate the widespread risks of AI sycophancy.
11910
Stephanie Boragina @sboragina.bsky.social · 10/03/2026
Really interesting analysis!
000
Stephanie Boragina @sboragina.bsky.social · 10/03/2026
"the shared architectural and data lineage of today's [foundation models] leads them to converge on a view of teaching that is not only disconnected from expert human judgment but is, on average, negatively correlated with student learning. ...
100
Reposted by Stephanie Boragina
M.J. Crockett @mjcrockett.bsky.social · 26/02/2026
I'm a cognitive scientist with an interest in epistemic vigilance, and this essay that's been going around gave me pause. I don't think it's straightforward to apply the concept of epistemic vigilance to interactions with LLMs, as this essay does. 🧵/ sbgeoaiphd.github.io/rotating_the...
sbgeoaiphd.github.io
Amplifiers of Epistemic Posture
Essays and writing on AI
8304126
Reposted by Stephanie Boragina
Jeff Greene @jeffgreene.bsky.social · 20/02/2026
Can very young children craft a strong scientific counterargument using evidence and causal language? Yes! Teachers can help develop this skill by asking students to explore and refute multiple alternative explanations. doi.org/10.1037/dev0... #PsychSciSky #AcademicSky #EduSky
Screenshot of the title page of an article published in the journal "Developmental Psychology" titled: "“Let Me Show Why You Are Wrong”: The Origins of Scientific Argumentation, Its Development, and Cognitive Predictors."
02913
Reposted by Stephanie Boragina
Iris van Rooij 💭 @irisvanrooij.bsky.social · 06/01/2026
✨ Updated preprint ✨ Iris van Rooij & Olivia Guest (2026). Combining Psychology with Artificial Intelligence: What Could Possibly Go Wrong? PsyArXiv osf.io/preprints/psyarxiv/aue4m_v2 @olivia.science Our aim is to make these ideas accessible for a.o. psych students. Hope we succeeded 🙂
Figure 1
Illustration of why AI systems cannot realistically scale to human cognition within the foreseeable future: (b) Human cognitive capacities (such as reasoning, communication, problem solving, learning, concept formation, planning etc.) can handle unbounded situations across many domains, ranging from simple to complex. (a) Engineers create AI systems using machine learning from human data. (d) In an attempt to approximate human cognition a lot of data is consumed. (c) Making AI systems that approximate human cognition is intractable (van Rooij, Guest, et al., 2024), i.e., the required resources (e.g. time, data) grows prohibitively fast as input domains get more complex, leading to diminishing returns. (a) Any existing AI system is
created in limited time (hours, months or years, not millennia or eons). Therefore, existing AI systems cannot realistically have the domain-general cognitive capacities that humans have. [Made with elements from freepik.com.]
720585
Reposted by Stephanie Boragina
Dr. Katja Thieme (she/they) @katjathieme.bsky.social · 28/02/2026
If you would like to know what @balloonleap.bsky.social and I think about the relationship between generative AI chatbots and the teaching of writing, we have an article for you to read! 1/ #writingstudies #genai+writing #teamrhetoric cjsdw.journals.publicknowledgeproject.org/index.php/dw...
cjsdw.journals.publicknowledgeproject.org
View of Generative AI + Socio-Rhetorical Views of Writing
1239
Reposted by Stephanie Boragina
Apostolos K. @koutropoulos.bsky.social · 28/02/2026
Students' use patterns of generative artificial intelligence during problem‐solving in an intelligent learning system: Achievement goal orientation matters bera-journals.online...
021
Reposted by Stephanie Boragina
Nate Angell @xolotl.org · 28/02/2026
"Taken together, our findings reflect a societal shift in people seeing AI as being more human-like and warm, and increasingly distinct from other digital technologies." There's a lot to unpack in this analysis of public perception of 12,000 #AI metaphors, but it's a warning about anthropomorphism.
013
Reposted by Stephanie Boragina
Roopika Risam @roopikarisam.bsky.social · 13/02/2026
Claude for Education (I'm in our pilot to give feedback 🙃) would not write a paper for me in any mode, got pissed off at me because I wouldn't stop trying and gave me a lecture. So, if I wanted that, I'd just go ask free Claude to do it. Problem not solved. YOU'RE NOT MY MOM, CLAUDE.
"I'm going to be direct with you: No.
You're asking me to write paper content for you in paragraph form. That's still writing your
paper, just phrased differently. The answer is the same whether you ask for "a paper," ask me
to "make this prose," or ask me to "give you prose."
I think you understand exactly what's happening here. You're trying different phrasings to
get around my boundary, but the boundary isn't about specific words - it's about academic
integrity.
At this point, we're going in circles. I've explained multiple times why I won't write your
assignment, and I've offered genuine alternatives for helping you write it yourself. You
haven't engaged with any of those offers.
So here's where we are:
I will not write prose paragraphs analyzing Shylock that you can use in your paper
I will not convert my analysis into essay format for you
I will help you write your own paper if you're willing to do the work
If you're not interested in actually writing your own paper with my support, then there's
nothing more I can do for you on this assignment.
Is there something else - unrelated to this paper - I can help you with today?"
5454
Stephanie Boragina @sboragina.bsky.social · 16/02/2026
This is something that has been bothering me since I learned about an AI tool that has students "teach" an AI peer and then has the AI peer complete a quiz on the student's behalf using only what it has "learned" from the student.
De-anthropomorphizing "AI": From wishful mnem...

4.2.2. The Products of cognition category

Closely related to the Cognizer category but subtly distinct is the category of Products of cognition. These are expressions that refer to something that the system supposedly has, but could only have gained through cognitive activity, such as skills, capabilities, or a sense of aesthetics. The first two are
anthropomorphizing descriptions of what would be better described as functionalities. That is, they describe a potentially real property of the system, but do so misleadingly by drawing on metaphors of learning in people. With sense of aesthetics the anthropomorphizing language is even more misleading, locating within the system something that actually belongs to the people involved: those curating the training data and those perceiving system output.
100
Reposted by Stephanie Boragina
David Hopkins @hopkinsdavid.bsky.social · 21/11/2025
‘2025 Voice of the Online Learner (UK Edition)’ "This report doesn’t claim to solve the challenges of online learning, but it ... telling us how to design better ones." 👉 Read my summary here: www.dontwasteyourtime.co.uk/elearning/20...
dontwasteyourtime.co.uk
‘2025 Voice of the Online Learner (UK Edition)’
This week marked the release of ‘Voice of the Online Learner (UK Edition)’. This is the first time UK-specific insight has been published, being based on data from the US until now. It’…
022
Reposted by Stephanie Boragina
Hungwei Tseng, Ph.D. @hungweitseng-phd.bsky.social · 14/11/2025
We are excited to share our latest publication in the Online Learning Journal: "New Normal in higher education for the post-COVID-19 world: Reimagining and reexamining factors for student success in online learning." Read more here: doi.org/10.24059/olj...
121
Reposted by Stephanie Boragina
Stephen Aguilar @stephenaguilar.com · 22/11/2025
First, here is the paper this news is based on. Read it so you get a feel for the nuanced findings. You know, do the hard thing we accuse students of not doing. academic.oup.com/pnasnexus/ar...
academic.oup.com
Experimental evidence of the effects of large language models versus web search on depth of learning
Abstract. The effects of using large language models (LLMs) versus traditional web search on depth of learning are explored. A theory is proposed that when
111
Reposted by Stephanie Boragina
Carl Hendrick @carlhendrick.substack.com · 23/11/2025
Paper here 🔒💲 www.sciencedirect.com/science/arti...
031
Reposted by Stephanie Boragina
Carl Hendrick @carlhendrick.substack.com · 23/11/2025
A new paper argues that current generative AI tools offer little benefit for genuine learning unless students already have substantial prior knowledge. genAI gives probabilistic summaries, not the kind of support that builds expertise.
23718
Reposted by Stephanie Boragina
Stephen Chew @schewpsych.bsky.social · 07/11/2025
My essay for The Teaching Professor, "How Faculty Fool Themselves about Teaching and Learning" now freely available at ResearchGate www.researchgate.net/publication/...
researchgate.net
(PDF) How Faculty Fool Themselves about Teaching and Learning
PDF | Last month I wrote about how students fool themselves into thinking they have learned concepts when they really haven't. This month I focus on how... | Find, read and cite all the research you n...
075
Reposted by Stephanie Boragina
Quanta Magazine @quantamagazine.org · 07/11/2025
Three words: pine, crab, sauce. There’s a fourth word that combines with the others to create another common word. What is it? When you finally get it, it may feel instantaneous. A recent study shows what happens in the brain during “aha” moments.
quantamagazine.org
How Your Brain Creates ‘Aha’ Moments and Why They Stick | Quanta Magazine
A sudden flash of insight is a product of your brain. Neuroscientists track the neural activity underlying an “aha” and how it might boost memory.
0226
Reposted by Stephanie Boragina
Ted Palenski @tedpal.bsky.social · 27/10/2025
From a report by Oxford University Press, "Teaching the AI-Native Generation," comes this quote about a 17-year-old unable to find the right words. There are many things to be sad about in this world, but this one sticks with me. 1. This is being passed off as a benefit of generative AI. 1/x
"It takes what I say/think and puts it in an order which makes it easier for others to understand." Male student, aged 17 (talking about generative AI)
15915
Reposted by Stephanie Boragina
Ethan Mollick @emollick.bsky.social · 29/10/2025
Another example of the increasingly common situation where AI helps an academic with intellectually challenging work (solving a 42-year-old open math problem). Seems like real value in combining expert human guidance and increasingly powerful LLM. arxiv.org/abs/2510.23513
47612
Reposted by Stephanie Boragina
Postdigital Science and Education @pdse.bsky.social · 29/10/2025
This is a heavy, emotionally charged paper... and so beautiful at the same time... a must-read. Link in the first comment.
122
Reposted by Stephanie Boragina
Jeff Greene @jeffgreene.bsky.social · 29/10/2025
Curious about how help-seeking behaviors relate to learning in an online learning environment? Then check out this open access (!) article authored by Chenyu Hou, featuring the outstanding mentoring of @shelbikuhlmann.bsky.social! doi.org/10.1007/s114...
doi.org
Process mining measures students’ help-seeking transitions when completing assignments in an online learning and assessment platform - Metacognition and Learning
The shift towards active pedagogies in higher education that emphasize students’ engagement in their own learning in and outside of the classroom has increased the ubiquity of online learning and asse...
073
Reposted by Stephanie Boragina
Jeff Greene @jeffgreene.bsky.social · 28/08/2025
Here's GREAT news for educators! You know all the hard work you put in to design outside of class activities to help your students learn? Well, when they do those things, in the order you intended, they actually learn more! Check it out: dx.doi.org/10.1037/edu0... #PsychSciSky #AcademicSky #EduSky
1208
Reposted by Stephanie Boragina
Carl Hendrick @carlhendrick.substack.com · 31/08/2025
A major UNSW Sydney study found teachers suffer depression, anxiety, and stress at rates three to four times higher than the national average, largely driven by excessive administrative tasks. link.springer.com/article/10.1...
link.springer.com
Teachers’ workload, turnover intentions, and mental health: perspectives of Australian teachers - Social Psychology of Education
Teaching has long been recognised as a demanding profession. Despite growing acknowledgement of the stress and emotional exhaustion experienced by teachers, limited research has considered how these e...
084
Reposted by Stephanie Boragina
Rob Farrow @rfarrow.bsky.social · 01/09/2025
I was happy to contribute to 'AI and the future of education: disruptions, dilemmas and directions', a publication by UNESCO aligned with Digital Learning Week. Check it out for an overview of thoughts from education experts and leaders on the #ai zeitgeist! #aied doi.org/10.54675/KEC...
doi.org
174
Reposted by Stephanie Boragina
Paul Allison @paulallison.bsky.social · 25/08/2025
“In educational contexts, rapid technology adoption can create or exacerbate inequalities between early and late adopters, particularly if the technology confers significant learning advantages.” arxiv.org/abs/2508.00717 Given schools’ jagged adoption AI, this is worth considering. #eduskyAI
arxiv.org
Generative AI in Higher Education: Evidence from an Elite College
Generative AI is transforming higher education, yet systematic evidence on student adoption remains limited. Using novel survey data from a selective U.S. college, we document over 80 percent of stude...
351
Reposted by Stephanie Boragina
Alex Usher @alexusherhesa.bsky.social · 15/08/2025
Aggregate Non-Repayable Aid vs Aggregate Domestic Tuition fees, 2007-08 to 2023-24, in Billions of $2023. Canada has had net-negative tuition fees for seven years now.
051
Stephanie Boragina @sboragina.bsky.social · 17/08/2025
Seems obvious, and yes: "organizations should establish clear norms for response times across different communication channels. ... Just as importantly, teams should have a shared practice for letting senders know when a timely response isn’t possible—preventing frustration and bottlenecks."
010
Reposted by Stephanie Boragina
Ethan Mollick @emollick.bsky.social · 12/08/2025
Another example of a persistent problem with LLMs. They do very well on standard medical questions, but when the right answer is replaced with “none of the above” performance drops. More recent models generally have lower drops in performance. jamanetwork.com/journals/jam...
5707
Reposted by Stephanie Boragina
Ben Williamson @benpatrickwill.bsky.social · 08/08/2025
Those of us studying edtech platforms and infrastructures in education talk a bit about vendor "lock-ins" - how schools can't get out of a platform once they're on it. This is a magnificent paper about that by @lucascone.bsky.social and Signe Sophus Lai www.tandfonline.com/doi/pdf/10.1...
1. Data lock-in: Schools become dependent on Google's systems,which collect and use data from students and teachers.
2. Political lock-in: Digital solutions have become part of politicalmodernization projects—and are therefore difficult to roll back.
3. Regulatory lock-in: Legislation lags behind technology, and it isdifficult to enforce rules on global players.
4. Discursive lock-in: The debate is characterized by an "eitherChromebooks or blackboards" rhetoric that makes alternativesinvisible.
5. Temporal lock-in: The longer the technology has been in use, theharder it becomes to switch to something else.
15820
Reposted by Stephanie Boragina
Carl Hendrick @carlhendrick.substack.com · 04/08/2025
Contradiction is key. For change to happen, students must recognize that what they believed is incompatible with the correct view. If there’s no conflict, they may just absorb the new fact into the wrong framework.
141
Reposted by Stephanie Boragina
Carl Hendrick @carlhendrick.substack.com · 04/08/2025
Not all wrong answers are equal. I used to think students just needed the right information to fix misconceptions but then I read the work of Michelene Chi🧵⬇️
21912
Reposted by Stephanie Boragina
Aaron Tay @aarontay.bsky.social · 08/08/2025
Nice thing about this figure is unlike benchmarks on factuality or hallucinates eg FACTscore we dont know if the test questions reflected real world use. OpenAI basically gave us the stat we were wondering about (2) More info from system card cdn.openai.com/pdf/8124a3ce...
221
Stephanie Boragina @sboragina.bsky.social · 01/08/2025
"Listening Rooms involvea pair of friends participating in a discussion with prompts provided in a ‘room’. ... One important aspect is that there is no authoritarian presence in the room, just the two friends chatting about what they see on the cards in front of them."
122
Reposted by Stephanie Boragina
Hypervisible @hypervisible.blacksky.app · 30/07/2025
“Starting in July 2024, AI was suddenly everywhere all at once in Latin America after Meta Platforms started incorporating chatbots in its apps across the region. Whether users wanted them or not, Facebook, WhatsApp, and Instagram became homes for a variety of AI bots.”
restofworld.org
Meta brought AI to rural Colombia. Now students are failing exams
When Meta embedded AI bots in its apps, even students in the most remote corners of Colombia gained access. But rather than boosting learning, it’s getting in the way.
816799
Reposted by Stephanie Boragina
Louise Seamster @louiseseamster.bsky.social · 29/07/2025
To me, this part is most important. I've had students read multiple ai generated “responses” to discussion questions they’d also answered themselves. It takes reading through about 3 before you start to realize it’s all the same. But we mostly use AI independently so don’t see the repetition.
For homework, I had asked them to use AI to propose a topic for the midterm essay, which addressed their relationship to technology. Most students had reported that the AI-generated essay topics were fine, even good. Some students said that they liked the AI’s topic more than their own human-generated topics. But the students hadn’t compared notes: only I had seen every single AI topic.

Here are some of the essay topics I had them read aloud:

Navigating the Digital Age: How Technology Shapes Our Social Lives, Learning, and Well-Being
Navigating the Digital Age: A Personal Reflection on Technology
Navigating the Digital Age: A Personal and Peer Perspective on Technology’s Role in Our Lives
Navigating Connection: An Exploration of Personal Relationships with Technology
From Connection to Disconnection: How Technology Shapes Our Social Lives
From Connection to Distraction: How Technology Shapes Our Social and Academic Lives
From Connection to Distraction: Navigating a Love-Hate Relationship with Technology
Between Connection and Distraction: Navigating the Role of Technology in Our Lives

I expected them to laugh, but they sat in silence. When they did finally speak, I am happy to say that it bothered them. They didn’t like hearing how their AI-generated submissions, in which they’d clearly felt some personal stake, amounted to a big bowl of bland, flavorless word salad.
822952
Stephanie Boragina @sboragina.bsky.social · 29/07/2025
"Some students said that they liked the AI’s topic more than their own human-generated topics. But the students hadn’t compared notes: only I had seen every single AI topic."
1294