Lightnews — Scholar-powered news

Reposted by Frederik Cornillie

Mr E @theniceman.bsky.social · 20d

Give Gemini AI a quote from a woman and it still attributes it to a man. Even though it could Google it. And in this case the man is much more obscure than the actual author.

Votre citation, "Malheureux celui qui veut gouverner autrui, et ne peut se gouverner lui-même !", est une profonde observation sur l'importance de l'autodiscipline et de l'intégrité personnelle pour quiconque aspire à un rôle de leadership.

Elle est souvent attribuée à l'écrivain et moraliste français Jacques-Bénigne Bossuet (1627-1704). Ce dernier, évêque de Meaux, était une figure majeure du XVIIe siècle, connu pour ses sermons et ses écrits sur la politique et la religion. Il soulignait souvent que la véritable autorité ne vient pas du pouvoir sur les autres, mais de la maîtrise de soi.

2

Reposted by Frederik Cornillie

CALICO - Computer Assisted Language Instruction Consortium @caliconsortium.bsky.social · 24d

📅 The deadline for submitting your abstract proposal will be November 2, 2025.

Learn about the beautiful Miami University campus and more details about the conference:

CALICO - Computer Assisted Language Instruction Consortium (@caliconsortium.bsky.social)

calico.org

bsky.app

2 1

Reposted by Frederik Cornillie

MIT Press @mitpress.bsky.social · 29d

"It is shameful that we know far more about the surface of Mars and its atmosphere than we know about parts of our oceans."

Gaia’s Fever: Why an Overheating Earth Could Turn Deadly

An aging Earth, like an aging body, is increasingly vulnerable to heat’s fatal accidents.

thereader.mitpress.mit.edu

7 20

Reposted by Frederik Cornillie

Neil Selwyn @neilselwyn.bsky.social · Jul 24

new paper 🔓! Teachers are having to put *considerable* amounts of work into making GenAI outputs suitable for their classrooms ... amidst the going hype we need to talk more about what GenAI clearly *cannot* do, and the uniqueness of teacher expert knowledge: www.tandfonline.com/doi/full/10....

When the prompting stops: exploring teachers’ work around the educational frailties of generative AI tools

Neil Selwyn, Marita Ljungqvist and Anders Sonesson [1]

Abstract: Teachers are now encouraged to use generative artificial intelligence (GenAI) tools to complete various school-related administrative tasks, with the promise of saving considerable amounts of time and effort. Drawing on interviews from 57 teachers across eight schools in Sweden and Australia, this paper explores teachers’ experiences when working with GenAI. In particular, it focuses on the large amounts of work that teachers put into reviewing, repairing and sometimes completely reworking AI-produced outputs that they perceive to be deficient. Rather than reflecting teachers’ lack of skill in prompting GenAI effectively, the paper shows how this work foregrounds the educational limitations and frailties of AI and other automated technologies – with teachers having to act on a wide range of complex professional judgements around pedagogical appropriateness, social relations and overall educational value to bear on AI-generated content. The paper concludes by considering the need to challenge ongoing industry and policy claims around the labour-saving benefits of artificial intelligence in education, and instead focus on the ways in which these technologies are dependent on the hidden labour of humans to co-produce the illusion of automation.

3 11 27

Reposted by Frederik Cornillie

Neil Selwyn @neilselwyn.bsky.social · Jul 14

Korean government looks set to give up on its much-heralded national 'AI textbook' policy after only a few months ... www.koreatimes.co.kr/southkorea/s...

AI textbooks targeted for phaseout under South Korea's new administration - The Korea Times

Following Korea’s presidential transition, significant changes are expected in education policy, as President Lee Jae-myung appears poised to halt the rollout of AI-powered digital textbooks. Once cha...

www.koreatimes.co.kr

1 4

Frederik Cornillie @fcornillie.bsky.social · Jul 10

Preview the research results bit.ly/3GyjASn , read more about the implementation of the learner model in the game bit.ly/3U2T6vi , or play with the adaptive system yourself, for free! www.languagehero.app . Full paper coming out this fall in CALICO Journal @calicojournal.bsky.social [3/3]

1

Frederik Cornillie @fcornillie.bsky.social · Jul 10

🔬 In our research with EdTech company Linguineo, we developed an efficient learner model for a game-based spoken dialogue system that is both transparent and grounded in theory about second language acquisition. And Linguineo has implemented the model in their game! [2/3]

1 1 4

Frederik Cornillie @fcornillie.bsky.social · Jul 10

🤖 LLM-based chatbots are increasingly being used as personal tutors for language learning, and will make you believe that they intelligently adapt to your proficiency. But adaptive systems for language teaching need to be transparent, in particular with regard to how they assess learning. [1/3]

1 3

Reposted by Frederik Cornillie

EuroSLA @eurosla.bsky.social · Jul 7

EuroSLA 35 will be in Lisbon, 24-27 June 2026

clunl.fcsh.unl.pt/en/eurosla-2...

EuroSLA 35 to be held in Lisbon in June 2026 - Centro de Linguística da Universidade Nova de Lisboa

The Research Center for Linguistics (CLUNL) at NOVA University Lisbon is proud to announce that the 35th Annual Conference of the European Second Language Association (EuroSLA 35) will take place in L...

clunl.fcsh.unl.pt

4 11

Frederik Cornillie @fcornillie.bsky.social · Jul 7

@nflrc.bsky.social

Frederik Cornillie @fcornillie.bsky.social · Jul 7

After 35 years of continuous activity with rigorous accountability, the network of 16 language resource centers in the US nflrc.org is being dismantled. If you are affected by or concerned about this, and are a US citizen, please contact your members of Congress via the link below!

1 1 2

Frederik Cornillie @fcornillie.bsky.social · Jul 3

Researchers in #CALL who were previously into games and play, and now seem to have disappeared onto the bandwagon of GenAI 🥱: you can still come back! 🙃 You can resist hype cycles! Slow research is fine!

3

Frederik Cornillie @fcornillie.bsky.social · Jul 3

More information: www.languagepolicy.org/post/jncl-nc...

JNCL-NCLIS Decries Massive Language Education Cuts in President’s Budget

On Friday, the White House Office of Management and Budget released a topline budget for Fiscal Year 2026 that would slash funding for many language education programs.

www.languagepolicy.org

1

Frederik Cornillie @fcornillie.bsky.social · Jul 3

Language education programs in the US are risking serious cuts, and these are imminent. If you are a US citizen, consider signing this petition: www.votervoice.net/Shares/BAAAA...

Language Education Funding in Jeopardy - Take Action Now

Language education funding for Fiscal Years 2025 and 2026 is in jeopardy. The Department of Education is months late in allocating to states the already enacted FY 25 funding for key language educatio...

www.votervoice.net

1 1 1

Frederik Cornillie @fcornillie.bsky.social · Jun 26

www.scientificamerican.com/article/repl...

Replacing Federal Workers with Chatbots Would Be a Dystopian Nightmare

The Trump administration sees an AI-driven federal workforce as more efficient. Instead, with chatbots unable to carry out critical tasks, it would be a diabolical mess

www.scientificamerican.com

Frederik Cornillie @fcornillie.bsky.social · Jun 23

I like getting out of my academic comfort zone and write with people who challenge my views. This collab with @york.bsky.social helped me rethink my position on games in language education. Games mustn't always be digital. Teachers are key. Tech can have other uses than teach language. Out now!

James York | ちーぷ @york.bsky.social · Jun 23

A chapter that I wrote with @fcornillie.bsky.social is finally out!

Teaching Languages with Games
👉 www.cambridge.org/core/books/c...

As with my other recent work, its published in a tech handbook, but we push for the inclusion of all digital and non-digital games under a #ludic umbrella.

Teaching Languages with Games (Chapter 10) - The Cambridge Handbook of Technology in Language Teaching and Learning

The Cambridge Handbook of Technology in Language Teaching and Learning - June 2025

www.cambridge.org

3

Frederik Cornillie @fcornillie.bsky.social · Jun 17

Very insightful read on the drawbacks of relying fully on current speech recognition systems like OpenAI's Whisper, without humans in the loop. Raises questions for education and in particular use cases that can have far-reaching consequences like high-stakes (language) testing. #EdTech #AI

Hypervisible @hypervisible.blacksky.app · Jun 17

“There is no ‘one weird trick’ that removes experts and creates miracle machines capable of doing everything humans can do but better.”

Replacing Federal Workers with Chatbots Would Be a Dystopian Nightmare

The Trump administration sees an AI-driven federal workforce as more efficient. Instead, with chatbots unable to carry out critical tasks, it would be a diabolical mess

www.scientificamerican.com

1 3

Reposted by Frederik Cornillie

CALICO - Computer Assisted Language Instruction Consortium @caliconsortium.bsky.social · Jun 12

CALICO has a new opportunity to contribute to research in our field. Check out the following opportunities in partnership with University of Toronto Press:
Chapter Proposals: bit.ly/4e17gX6
Book Proposals: bit.ly/3FV1ols

CFP 2027 - Advances in CALL - CALL and the intersection between language and culture

Dear colleagues, We are currently seeking chapter proposals for the 2027 Advances in CALL Research and Practice edited volume. In the volume, entitled CALL at the intersection between language and cu...

bit.ly

1

Frederik Cornillie @fcornillie.bsky.social · Jun 7

"De relatie tussen kennis en welbevinden: tweerichtingsverkeer"

duurzaamonderwijs.com/2025/05/27/d...

#evidence-based #onderwijs

De relatie tussen kennis en welbevinden: tweerichtingsverkeer

Als kinderen zich goed voelen op school, kan dat de mate waarin ze op school nieuwe kennis verwerven positief beïnvloeden. Omgekeerd kan het succesvol verwerven van nieuwe kennis het welbevinden va…

duurzaamonderwijs.com

Reposted by Frederik Cornillie

FanTALES @fantales.bsky.social · May 23

We have a new banner (long overdue 😷) for our website. 😁Let us know what you think !

1

Frederik Cornillie @fcornillie.bsky.social · May 9

One of my favorite events at CALICO Conference! #startups #edtech #languagelearning

CALICO - Computer Assisted Language Instruction Consortium @caliconsortium.bsky.social · May 8

Join us for the LaunchPad competition at #CALICO2025 !

The live pitch event showcases 3 groundbreaking language learning startups.

🕦 Fri May 30, 11:30–12:45pm

📍 SDSU Aztec Union Theatre

Don't miss one of the conference highlights for EdTech & CALL innovation
thelanguageflagship.tech/launchpad/

LaunchPad – Int’l Language Education Tech Competition

A lifetime opportunity for startups to showcase their innovation to an audience interested in foreign language learning or teaching.

thelanguageflagship.tech

1

Reposted by Frederik Cornillie

CALICO - Computer Assisted Language Instruction Consortium @caliconsortium.bsky.social · May 8

Join us for the LaunchPad competition at #CALICO2025 !

The live pitch event showcases 3 groundbreaking language learning startups.

🕦 Fri May 30, 11:30–12:45pm

📍 SDSU Aztec Union Theatre

Don't miss one of the conference highlights for EdTech & CALL innovation
thelanguageflagship.tech/launchpad/

LaunchPad – Int’l Language Education Tech Competition

A lifetime opportunity for startups to showcase their innovation to an audience interested in foreign language learning or teaching.

thelanguageflagship.tech

2 2

Reposted by Frederik Cornillie

Emily M. Bender @emilymbender.bsky.social · Apr 24

LLMs are nothing more than models of the distribution of the word forms in their training data, with weights modified by post-training to produce somewhat different distributions. Unless your use case requires a model of a distribution of word forms in text, indeed, they suck and aren't useful.

Hank Green @hankgreen.bsky.social · Apr 24

There are a lot of critiques of LLMs that I agree with but "they suck and aren't useful" doesn't really hold water.

I understand people not using them because of social, economic, and environmental concerns. And I also understand people using them because they can be very useful.

Thoughts?

55 490 1.6K

Reposted by Frederik Cornillie

Shannon Sauro @shansauro.bsky.social · Apr 24

Whenever I encounter scholars extoling the benefits of AI for fans, I can’t help but wonder what kind of fans they are talking about because it’s absolutely not the fanfic writers on the fanfic archive AO3. This includes me, btw.

Elizabeth Minkel @elizabethminkel.bsky.social · Apr 24

🫠🫠🫠

www.reddit.com/r/AO3/commen...

AO3 has been scraped. Again. For GenAI purposes.

www.reddit.com

1 1 5

Reposted by Frederik Cornillie

Jon Ingold @jon.inkle.co · Apr 17

List piece on dualshockers about linguistic games - and great, HV is in it, but also - PataNoir! The Gostak! Suveh Nux! This is a good list.

The Gostak is an all-timer.

8 Best Games For Linguistic Experts

The best video games for players than consider themselves linguistic aficionados.

www.dualshockers.com

1 21 64