Ben Williamson

@benpatrickwill

Researching data, tech, futures, and biological sciences in education | Senior Lecturer and co-director at the Centre for Research in Digital Education | University of Edinburgh | Editor of Learning, Media and Technology @lmt-journal.bsky.social

1550 Siguiendo7855 Seguidores1570 Publicaciones

Bluesky

¿Es tu perfil?Reclámalo y benpatrickwill.dirty.social es tuyo.Reclamar perfil

Ben Williamson@benpatrickwill· 23h

As if populating a course on Blackboard wasn't bad enough, now it features a gallery of "automations" you can add. This celebratory feedback for "outstanding achievement" is both gross but also comically naff. "Your potential is limitless"🤣Imagine addressing your students like this in person!

Screenshot of an automated message template on the Blackboard LMS interface, with text reading:

From:
Ben Williamson

Instructor
Message:
I wanted to take a moment to celebrate your outstanding achievement.

Your hard work, dedication, and focus have truly paid off, and I couldn't be more proud of you!

Achieving such an exemplary grade is no small feat, and it's a clear reflection of your commitment to excellence.

Keep up the amazing work – this is just the beginning of many more successes to come.

Continue to aim high, and remember that your potential is limitless!

Wishing you continued growth and success in all that you do.

Well done once again!

Ben Williamson@benpatrickwill· 5d

Looks like if you're one of the 300miliion who ever took a MOOC with Coursera or Udemy, your data just got turned into assets for Andrew Ng's latest "AI native" edtech "personalized learning" scheme, which Coursera just invested $100m in. investor.coursera.com/news/news-de...

Our strategic investment in LearnVector has the potential to be a force multiplier for our growth. Andrew's unmatched AI expertise pairs his agentic AI with our platform and assets that turn learning into validated mastery — a clear competitive advantage that we believe can accelerate our business and expand the impact we deliver for individuals and enterprises around the world.
— Greg Hart, CEO, Coursera

Source https://learnvector.ai/

Building on the broader reach we’ve gained through Udemy, this investment reflects our conviction that AI can help grow the market itself by making learning more continuous, more personal, and more embedded in people’s daily lives. In other words, we’re focused on creating a much bigger opportunity for learning overall.

As LearnVector develops new AI-native learning experiences, we’re exploring powerful ways to combine those advances with Coursera’s unique strengths: trusted content, insights from more than 300 million learners, and one of the world’s largest ecosystems of university and industry partners.

Source https://blog.coursera.org/coursera-invests-in-learnvector-to-build-the-future-of-ai-native-learning/

Ben Williamson@benpatrickwill· Jun 30

My chapter is a small contribution to a really cool book project by Sarah Truman "Research and Teaching with Speculative Fiction: Transdisciplinary Readings and Methods" - the whole book is OA with short accessibke chapters www.bloomsburycollections.com/monograph?do...

Cover image of Research and Teaching with Speculative Fiction: Transdisciplinary Readings and Methods

Ben Williamson@benpatrickwill· Jun 11

New chapter on the scientific infrastructure-building that took place in educational genomics and neuroscience over the last two decades. The chapter is a synoptic version of our work on the "bio-edu-data sciences" funded by @leverhulme.ac.uk and is now out (paywalled) here.. doi.org/10.4337/9781...

Recent advances in educational neuroscience and educational genomics have been made possible by the incorporation of advanced technologies into scientific research designs. Both educational genomics and neuroscience have become data-intensive biological enterprises. In this chapter, I unpack the emerging scientific infrastructures underpinning these ‘bio-edu-data sciences’. The analysis foregrounds (1) infrastructuring: the creation of novel sociotechnical systems of investigation and knowledge production; (2) informatting: the transformation of molecular biological substances into informational formats; and (3) intervening: the translation of bioinformation considered relevant to education into proposals for interventions in policy and practice. The bio-edu-data sciences represent a paradigm shift in research on the determinants of educational outcomes, shifting research practices into data-scientific laboratories that resemble those of ‘big biology’ programmes, prioritizing the analysis of very large volumes of bioinformation and directing the attention of scientists, educators, and policymakers to the molecular substrates of learning achievements.

Ben Williamson@benpatrickwill· May 8

Every time Canvas is in the news I'm reminded of this excited statement about its massive stores of "data assets" from its ceo in 2019 eliterate.us/instructure-...

We have the most comprehensive database on the educational experience in the globe. So given that information that we have, no one else has those data assets at their fingertips to be able to develop those algorithms and predictive models.

Ben Williamson@benpatrickwill· Mar 29

New article just out exploring how "educational genomics" scientists propose using polygenic scores as the basis for policy and practice interventions, and what this geneticization of education entails. doi.org/10.1177/1474...

The use of genetic data has proliferated in social and behavioural genomics research, leading to proposals for genetically-informed social policy. In particular, in the domain of ‘educational genomics’, researchers using ‘polygenic scores’—summary statistics used to calculate social outcomes from genetic data—have generated scientific knowledge claims about the potential to predict educational outcomes from biological samples. These polygenic scores have become the basis for proposed experimental interventions in education policy and practice. In this article analyzing a large corpus of scientific texts, we develop an original conceptualization of five techniques associated with the use of polygenic scores in education. We identify the underlying scientific base for claims of the policy-relevance of educational genomics, and develop a conceptual vocabulary explaining the implications and significance of polygenic scores in education. We conceive of polygenic scores as bio-social policy instruments that fuse biological knowledge to social policy objectives. Educational genomics represents an experiment in genomic social policy, amplifying policy commitments to rating, sorting, evaluating, stratifying, and valuing students in education through promissory claims of biological authority. Polygenic scores have thus become the basis for educational genomics to exert policy influence as a significant and controversial source of educational intervention and governance.

Ben Williamson@benpatrickwill· Mar 27

On "AI Literacy Day" I suggest reading "The (im)possibility of AI literacy" by @lucipangrazio.bsky.social questioning whether "literacy" is even the right response to AI and, if so, how a meaningful AI literacy could build on the history of "critical digital literacies" doi.org/10.1080/1743...

So what kind of response to AI do we need? Rather than just turning to literacy for the answers, we need to carefully consider what kind of ‘text’ AI is. Is it amenable to a literacies response? The stakes are high if we do not think carefully about an appropriate educative response. Not only will we not use AI effectively, ethically or well, we will stop looking for a more suitable and perhaps more robust response to it. We also risk literacy being coopted for compliance and productivity purposes, so it operates as a kind of ‘soft governance’ for participation in the digital economy (Pangrazio and Sefton-Green Citation2024), just as it did in the late nineteenth century when it was used to teach values and morality.

If literacy is the right response, then it needs to be more nuanced in how it is operationalised. Currently, it is used in ways that are both too narrow to capture the digital platforms and political and economic systems it is embedded in, but also too broad to capture the huge variations of how it is employed.

Ben Williamson@benpatrickwill· Mar 14

Scienmag is a bit of a treasure trove of both AI-made imagery and text - it's like fully automated SciComms of the most embarrassing kind

Image from a Scienmag article depicting a baby's face with glowing red protuberances all over it

Ben Williamson@benpatrickwill· Mar 13

Won't link to it but this is the thing "Scienmag" and the others do

Metrics page of article titled "Emotional Artificial Intelligence in Education: A Systematic Review and Meta-Analysis" with "9688 accesses"

Ben Williamson@benpatrickwill· Mar 6

Shouldn't let OpenAI puff about education get to me, but its claim students need AI skills for future jobs and "agency", then letting slip it really means making students into GPT "power users", and making that an educational imperative, *and* planning to measure it ... 😦 openai.com/index/ai-edu...

Text from OpenAI article on its educational programs:

"Agency does not emerge from basic AI use alone. Students must progress from simple tasks to deeper applications such as studying, building, creating, coding, and managing agents. But among college-age users, we see a widening global “capability overhang,” defined as the gap between what AI tools can do and how people actually use them. Even advanced student users still operate roughly 90% to 99% below how power users of ChatGPT are engaging with our tools.

Educational institutions play a central role in closing that gap. Faculty and educators can help students harness AI’s full potential by embedding authentic AI use cases into coursework—assignments that use AI and mirror real professional work, such as analyzing a market, designing a product concept, evaluating a policy trade-off, or building a simple agent"

Ben Williamson@benpatrickwill· Feb 27

Private equity is now a key actor - discursively, organizationally, financially - in shaping the education sector. But its gaze sees only future returns from financial calculations on asset control. That's not how most education workers see the value of the sector www.macfarlanes.com/insights/102...

Text reads:

Conclusion
The education sector presents a compelling investment opportunity shaped by geographic concentration in an evolving market. Education's unique positioning as economic infrastructure underpins its investment appeal. The sector's role as a GDP multiplier, where each 1% increase in education expenditure corresponds to 21.3% GDP growth, distinguishes it from traditional service sectors through both direct returns and broader economic value creation. This dual nature provides defensive characteristics whilst creating substantial consolidation opportunities across a market projected to reach $8tn by 2030.

Ben Williamson@benpatrickwill· Feb 27

Third, from a private equity perspective, public sector austerity is an opportunity for massively scalable private technosolutions to be rolled out at pace. It's a common narrative of the investor as a "saviour" of the public sector's failings...

Text reads:

Public sector funding constraints present significant opportunities for private capital to drive technology-enabled solutions that address growing skills gaps6 whilst delivering more scalable, cost-effective, and interactive educational experiences. Unlike public institutions constrained by budget cycles and procurement processes, private capital can move quickly to implement technological solutions and respond to evolving market needs.7

Ben Williamson@benpatrickwill· Feb 27

Second, private equity views education as "human capital" investment. So education is configured as *valuable* from an investors' vantage point in terms of "positive externalities" and "spillover effects" into "productivity". This feels like an image of education that AI would amplify...

Text reads:

Education's role in human capital development creates sustainable competitive advantages that transcend economic cycles. Research shows that each extra year of schooling leads to an approximately 9% annual increase in a person’s earnings,and the associated human capital generates positive externalities and spillover effects that benefit entire economies through enhanced productivity and innovation.5 The sector’s inelastic demand also provides defensive investment qualities particularly valuable during periods of economic uncertainty.

Ben Williamson@benpatrickwill· Feb 27

How the private equity industry views education is really alien to how educators see it. Yet these investors probably have a bigger say in how the future of education will unfold. First, education is a *market* worth $trillions *in the future* making it a good speculative bet for investment...

Text reads:

Why education?
Education’s investment appeal stems from its fundamental role as economic infrastructure. The global education market is projected to expand from $6tn in 2022 to $8tn by 20303, representing a compound annual growth rate (CAGR) of 3.7%, a trajectory that reflects sustained market expansion despite economic uncertainties.

Ben Williamson@benpatrickwill· Feb 20

Elsevier, Wiley, then the rest will follow, turning academic labour into assets with constant financial returns into the future through their alliances with AI firms and the data licensing agreements they made to provide them our work for LLM training.

Screenshot reads:

Elsevier isn’t alone in staking its claim in this burgeoning field. Wiley, another large publisher, is teaming with publishers to develop AI Gateway, a platform that allows users to employ any of several LLM for content analysis.

Ben Williamson@benpatrickwill· Feb 20

The inevitable next stage of academic publishers profiting from academics' work is here - scraping it for AI then charging subscriptions for access to the AI summaries, and then again for the citations. Academic content assetization as we called it in a recent paper. www.science.org/content/arti...

Screenshot reads:

She and other publishing specialists question whether LeapSpace’s limited reach is worth the cost. Users will need either an institutional subscription (based in part on the institution’s size and amount of research) or an individual one, which costs $32 a month. Many libraries are already struggling to afford existing subscriptions. And if users want to read the cited content, they will need a separate subscription to that content’s publisher—akin to paying for multiple video-streaming services.

Ben Williamson@benpatrickwill· Feb 14

The "AI revolution" in science is going so smoothly that my instititutional library is having to deploy multiple technical protection measures against "poorly behaved Gen AI crawling" just to maintain its timely service library.ed.ac.uk/research-sup...

Screenshot of text from University of Edinburgh library services:

Like many academic libraries, we have been experiencing slowness and outages on some of our open access platforms over the last few months. We have identified that this is the result of poorly behaved Gen AI crawling activity. As such we have had to try to disable the ability for AI -harvesters to access content from our repositories, via IP range blocking, geographical blacklisting, implementing limits to throttle bots and using robots.txt files to limit access.

We acknowledge that this approach is not consistent with our open access policy which should require us to permit machine access to our content. Our limiting activity is not 100% fool-proof, and is a temporary measure to give services breathing space to perform while a longer termn solution is worked out by software providers and the community. This approach also has the added benefit of helping to assuage authors fear of losing control of their works.

Ben Williamson@benpatrickwill· Jan 30

How can a paper that was never written end up with 60+ citations in multiple journals and books with several different author and co-author credits? Maybe it started with human error/malpractice, but AI is clearly reproducing and amplifying the zombie infection.

Screenshot of a list of "zombie citations" with corresponding texts citing them.

Ben Williamson@benpatrickwill· Jan 30

When I posted before about finding a zombie reference, it had turned up on Google Scholar with 40ish citations as an undead paper. Now Scholar shows multiple citation stubs and escalating citation numbers despite it still not actually existing. It's got a half-life of its own.

Screenshot from Google Scholar of 3 citation stubs for a publication titled "Education governance and datafication".

Ben Williamson@benpatrickwill· Jan 28

JFC "Andrew" the "AI-generated expert on artificial intelligence" on this Microsoft course is the worst teacher and my already dim view of AI in education and training has plummeted even further now.

Screenshot of "Andrew", an AI-generated trainer delivering a Microsoft-produced course about AI concepts

Ben Williamson@benpatrickwill· Jan 28

The thread was right though - the first course I clicked on (Microsoft "Introduction to AI concepts") is led by an AI-generated pedagogic avatar and the first review is not glowing

Screenshot of a review of a course on the AI Skills Boost Hub, reading: "A pretty good introduction to the concepts. However, I find it incredible that an AI generated video that extols the virtues of synthesised speech does not give me the choice of listening in my own language - not American English."

Ben Williamson@benpatrickwill· Jan 28

I signed up to the government's AI Skills Boost Hub to check its suite of courses and I am sad to report that the free AI training looks even worse than the thread below suggests aiskillshub.org.uk/aiskillsboost/

Screenshot from the AI Skills Boost Hub website launched by the Department for Science, Innovation and Technology, showing courses offered by Microsoft.

Ben Williamson@benpatrickwill· Jan 24

When Anthropic says it's recruiting 100,000 teachers to be "co-architects" of AI in education through partnering with the global Teach for All network, that sounds like a lot of free labour, brand advocacy, and integration into teaching that it'll benefit from www.anthropic.com/news/anthrop...

Screenshot from Anthropic press release announcing partnership with Teach for All:

"Built on shared learning
What makes this partnership distinctive is its approach. Teachers are positioned not as passive consumers of AI tools, but as co-architects shaping how AI develops. Through the AI LCC, Anthropic provides access to Claude, while educators provide on-the-ground feedback to inform how the product evolves.

“For AI to reach its potential to make education more equitable, teachers need to be the ones shaping how it's used and providing input on how it's designed,” said CEO of Teach For All Wendy Kopp. “Our partnership with Anthropic is helping educators across our network experiment with and learn from these tools firsthand, as co-creators of AI's role in education.”"

Ben Williamson@benpatrickwill· Jan 19

Wild to me that the Department for Education has pushed generative AI for use in schools but now imposing a bunch of new standards that amount to requiring surveillance of students to ensure they're not using the tech for "cognitive offloading" or "emotional engagement" www.gov.uk/government/p...

Screenshot of text from Department for Education standards for generative AI products:

"As stated in the relevant sections of these standards, products should monitor, regularly report on, and provide data to teachers on:

the rate of requests for cognitive offloading and the amount of cognitive offloading delivered

the level of personal and emotional engagement by each user in terms of the nature of information exchanged, without directly disclosing the content of these inputs

the duration of usage by each individual learner"

Ben Williamson@benpatrickwill· Dec 23

It even appears as a *book* with an extra subtitle in one of them! Just incredible really. Thanks for deciphering what's going on

Ben Williamson@benpatrickwill· Dec 13

Learning management systems are "platforms" and they follow the platform dual income model of subscription contracts plus data extraction. *Our* data are *their* assets, which they can capitalize however they wish due to the terms of service our institutions sign www.ei-ie.org/en/item/2848...

The Instructure company highlights how edtech platforms’ data assets may be subject to unanticipated future uses. These potential uses are covered in their legal policies:
Instructure may share information in any talks about change of control: ‘We may share information about you in connection with or during negotiation of any merger, financing, acquisition, bankruptcy,
dissolution, transaction orproceeding involving sale, transfer, divestiture or disclosure of all or a portion of our business or assets to another company. (European Union Region Product Privacy Notice
Addendum)
In this respect, user data is increasingly considered as a platform company’s assets. Aligned with the dynamic business approaches of the digital economy more broadly, this means data are often collected for the potential value they might yield in the future, not necessarily because of the current requirements of a platform or for wider educational purposes.

Ben Williamson@benpatrickwill· Dec 6

Researcher produces 100+ papers on "AI" in a year and what's worse than the shonky vibe scholarship and "slop flood" this represents is that he's doing it all as boss of an edtech company called Algoverse charging students $$thousands to do the actual work www.theguardian.com/technology/2...

"In response to a query from the Guardian, Zhu said that he had supervised the 131 papers, which were “team endeavors” run by his company, Algoverse. The company charges $3,325 to high-school students and undergraduates for a selective 12-week online mentoring experience – which involves help submitting work to conferences.

“At a minimum, I help review methodology and experimental design in proposals, and I read and comment on full paper drafts before submission,” he said, adding that projects on subjects such as linguistics, healthcare or education involved “principal investigators or mentors with relevant expertise”."

Ben Williamson@benpatrickwill· Nov 29

Edtech has absolutely turned the language of regulation to its own advantage and maximization of opportunities. Also it is depressing that actual people utter words in this way now.

Screenshot reading:

Gesess also illustrates how regulatory shifts can spark innovation. When Italy altered the medical school entry exam, Futura, his AI tutoring platform, pivoted from test prep to ongoing, personalised tutoring. Turning policy disruption into opportunity is vital for founders today.

Rahim Hirji, EdTech strategist and author of SuperSkills, adds, “Europe’s regulatory framework recognises this truth. By insisting on transparency, human oversight, and accountability from the beginning, it protects the integrity of learning itself. It represents the architecture of trust on which lasting innovation depends. The result is an EdTech model built for human growth rather than machine efficiency, giving Europe a quiet but lasting advantage as the world decides what kind of learning future it wants.”

Ben Williamson@benpatrickwill· Nov 28

Edtech companies and their venture capital investors are not ignorant about regulatory and ethical regimes - they are *playing* them as a "growth lever" to conduct student surveillance and data extraction techfundingnews.com/expert-talk-...

Screenshot reading:

The investment thesis for regulation-literate founders

For venture capitalists, Marovac sets out a clear thesis: “Europe’s regulatory environment requires patience and a long-term view but rewards discipline. Regulation isn’t just a risk — it’s a catalyst for innovation. The best European EdTech startups harness it as a growth lever.”

Ben Williamson@benpatrickwill· Nov 22

Feeling queasy now I've learned that Google is not only now offering Gemini ambassador certificates to teachers but also awarding students and *children* to be corporate AI mascots. Coercive normalization of commercialized digital education. blog.google/outreach-ini...