Sign in

Nabil Hathout

@nabilhathout.bsky.social
164 followers 375 following 41 posts

Morphology, Paradigms, NLP, Language resources CNRS, CLLE, Toulouse

PostsRepliesMedia
Reposted by Nabil Hathout
Daria Dayter @coocho.bsky.social · 20/09/2026
The idea that someone somehow would think that this system is superior to basic funding… is mad. how on earth did we get to this place
2399
Reposted by Nabil Hathout
Julien Gossa @juliengossa.cpesr.fr · 01/09/2026
[ #VeilleESR #Obscurantisme ] La NSF (ANR US) réduit les financements pour les sciences sociales dans les programmes prestigieux. Après avoir abandonné l'histoire et philosophie des sciences, suivent : science politique, politique publique, anthropologie culturelle et enseignement des sciences.
science.org
NSF shrinks support for social sciences in prestigious graduate fellowship program
New solicitation excludes students in seven fields and science education
21312
Reposted by Nabil Hathout
Brent Toderian @brenttoderian.bsky.social · 29/08/2026
“Dutch people cycle an average of 2.6km each per day. If this pattern was replicated worldwide, annual carbon emissions would drop by 686 million tonnes. This mammoth figure exceeds the entire carbon footprint of most countries, including the UK, Canada, Saudi Arabia and Australia.” @euronews.com
euronews.com
Cycling like the Dutch would slash the world’s carbon footprint
If everybody cycled like the Dutch, we could offset the UK or Australia’s entire carbon footprint.
741775695
Reposted by Nabil Hathout
Julien Gossa @juliengossa.cpesr.fr · 29/08/2026
[ #VeilleESR #Parcoursup ] Après le classement des meilleurs formations, le classement des formations privées en liquidation qui laissent leurs étudiants endettés sur le carreau. Félicitons encore une fois notre glorieux président et sa glorieuse ministre du travail pour cette formidable réussite.
75secondes.fr
Tsunami dans l'éducation : ces 50 écoles privées et CFA qui s'effondrent en silence
Faillites en série, étudiants piégés et profs impayés : découvrez la liste noire explosive des écoles privées et CFA liquidés en 2026.
19331230
Reposted by Nabil Hathout
Kevin Cariou @cariouk.bsky.social · 27/08/2026
Je vous dirais bien qu'on ne mesure pas la science au kilo de de publication et que plein de ces métriques sont bof bof... Mais il faut bien reconnaître que quel que soit le bout de la lorgnette par lequel on regarde, on ne peut que constater que la Science en France se porte de plus en plus mal.
1129
Reposted by Nabil Hathout
Libération @liberation.fr · 27/08/2026
La science française en décrochage : «Les faits sont là, l’érosion est bien réelle» ➡️ En savoir plus : l.liberation.fr/BtT
1012668
Reposted by Nabil Hathout
Maria Antoniak @mariaa.bsky.social · 28/08/2026
Truly, end of an era.
0345
Reposted by Nabil Hathout
Dr Michel Bessiere @michelbessiere.cpesr.fr · 26/08/2026
233
Reposted by Nabil Hathout
Thomas Pellard @thomaspellard.bsky.social · 18/08/2026
New preprint! (Yes, another one) Exploring the impact of inference method, character coding scheme, and cognacy criterion on phylolinguistic inference: The case of the Japonic language family osf.io/preprints/so...
osf.io
OSF
053
Reposted by Nabil Hathout
Thomas Pellard @thomaspellard.bsky.social · 18/08/2026
New preprint!Japanese has effectively only two accent patterns: A quantitative and information-theoretic approach to lexical predictability osf.io/preprints/so...
osf.io
OSF
074
Reposted by Nabil Hathout
Katie Mack @astrokatie.com · 17/08/2026
14 hours later and my mentions are still full of angry treatises about badly behaved cyclists. Bike lanes are not a reward for good cyclist behavior! They are infrastructure proven to make cities safer for ALL ROAD USERS. Whatever way cyclists have wronged you is irrelevant to the discussion!
303109261845
Reposted by Nabil Hathout
Jeffrey Brainard @jeffreybrainard.bsky.social · 12/08/2026
Hundreds of ads by #papermills, offering authorship on scientific articles for a fee, wound up published in conference proceedings. Are these venues a back-door route that’s too easy? #scicomm #retractions #peerreview @science.org www.science.org/content/arti...
science.org
Hundreds of paper-mill papers peddled in ads were later published
More than one-fifth of some IEEE conferences’ papers appear linked to authorship for sale
11914
Reposted by Nabil Hathout
David Monniaux @monniauxd.bsky.social · 13/08/2026
Régulièrement des collègues de LSHS semblent découvrir que Cairn est un site commercial sans ligne scientifique !
1162
Nabil Hathout @nabilhathout.bsky.social · 13/08/2026
New work on how to transpose the Paradigm Cell Filling Problem to derivation. Work done with Basilio Calderone, Franck Sajous and Fiammetta Namer, to appear in Word Structure (EUP). hal.science/hal-05714785
Abstract. A computational exploration of the Paradigm Cell Filling Problem in derivation
141
Reposted by Nabil Hathout
Nature @nature.com · 11/08/2026
A new wave of grant funders is turning to lotteries to break ties between proposals go.nature.com/4wNlgw1
go.nature.com
When the best decision is no decision: the rise of randomization in grant funding
If reviewers can’t reliably tell the difference between top grant proposals, why should they pretend they can? A new wave of funders is turning to lotteries to break the tie.
12519
Reposted by Nabil Hathout
John Bike Lane @johnbikelane.eurosky.social · 06/08/2026
1247
Reposted by Nabil Hathout
FX Coudert @nanoporous.net · 06/08/2026
In case you missed it: @chemicalscience.rsc.org now charges an APC of £2,850 per paper. It was previously free (diamond OA), and in my view it contributed greatly to the appeal and success of Chem Sci. Sad to see the RSC go that way…
For submissions from 1 July 2026 an APC will apply to all articles accepted in Chemical Science. Use our journal finder to check if your APC is covered by an institutional agreement with us. Note, if your institution has an agreement with us that covers APCs, you will be notified of this at article acceptance. Details of the APC and discounted rates are found here.

Get your manuscript submission ready with our submission checker tool. You can instantly check against Chemical Science's guidelines and requirements, and receive helpful hints to get you ready to submit.
2179
Reposted by Nabil Hathout
Thomas Steinke @stein.ke · 04/08/2026
My radical suggestion for peer review: Move from nominal pre-publication review to explicit post-publication review. The original reason for peer review was that journal pages are a limited resource, so we need a filter before publication. That no longer makes sense with digital publishing. 1/
2215
Reposted by Nabil Hathout
EACL 2027 @eaclmeeting.bsky.social · 04/08/2026
4134 submissions have been submitted to the ARR August cycle, a 14% increase compared to the ARR October 2025 cycle corresponding to EACL🔥. #EACL2027 in Athens is likely to be the biggest EACL ever 👀 #NLProc @aclrollingreview.bsky.social
01712
Reposted by Nabil Hathout
Léo Varnet @leovarnet.bsky.social · 01/08/2026
"Trois ans de recherches de financements", quelques conclusions tirées de mes échecs aux différents appels à projets. dbao.leo-varnet.fr/2026/08/01/t...
The House of Dust, Alison Knowles. Une imprimante qui génère des pages et des pages de poésie automatique.
121
Reposted by Nabil Hathout
Gautam Kamath @gautamkamath.com · 30/07/2026
ICLR 2027 has authorship quotas: - No more than 20 submissions per author - No more than 1 submission where no author has a paper previously accepted to a major ML conference From iclr.cc/Conferences/...
54813
Reposted by Nabil Hathout
Mehdi Senoussi @cogsenoussi.bsky.social · 31/07/2026
📢 PhD opening in comp. cog. neuro. in Toulouse: thalamo-frontal model of cognitive control. Modelling/EEG experience is welcome but not required. Application deadline: 8 Sept. 2026 Details: sdrive.cnrs.fr/s/eDrkeXfjdW... #CognitiveScience #Neuroscience #EEG #neuroskyence #compneurosky
02527
Reposted by Nabil Hathout
Markˣ @shjsat.bsky.social · 30/07/2026
I'm cross-posting @haspelmath.bsky.social "Gradient in grammatical structure of indigenous languages reflects pathway of human expansion in the Americas" Link to paper: www.nature.com/articles/s41...
Facebook post by Martin Haspelmath discussing Urban & Guzman-Naranjo's paper
124
Reposted by Nabil Hathout
Leo Hickman @leohickman.carbonbrief.org · 30/07/2026
++BREAKING++ The UK and Spanish governments have issued a joint statement overnight: "[We] agreed that this summer’s wildfires demonstrated that climate change was now a national security emergency facing Europe and threatening our way of life." www.gov.uk/government/n...
11332511369
Reposted by Nabil Hathout
Julien Gossa @juliengossa.cpesr.fr · 27/07/2026
L'article n'indique pas sous quel statut les chercheurs deviendraient des personnels des universités. Reste un indice fort en faveur de mon hypothèse « Fusionner les chercheurs et les enseignants-chercheurs permet de rendre acceptable l’acte 2 de l’autonomie ».
blog.educpros.fr
La guerre des tout petits mondes : bataille autour de l’acte 2 de l’autonomie
Sous pression budgétaire et politique, dépourvus de projet stratégique enthousiasmant, peinant à s’adapter au XXIe siècle, et privés des arbitrages d’un ministère trop occupé par les changement de …
1105
Reposted by Nabil Hathout
Gro-Tsen @gro-tsen.bsky.social · 23/07/2026
… journals and preprint servers being drowned in AI slop will probably result in any paper submission needing to come from someone with a verifiable academic position, or vouched by someone who has one, rather than on the paper's merits. This is very bad. •3/7
2102
Reposted by Nabil Hathout
Julien Gossa @juliengossa.cpesr.fr · 23/07/2026
The government said the move “will align our science and research base directly with commercial application and job creation” "it is increasingly clear that the sector as it has existed for a generation is unsustainable”
053
Reposted by Nabil Hathout
bat @batgo.bsky.social · 23/07/2026
Bientôt aussi près de chez vous
22111
Reposted by Nabil Hathout
Thomas Jj @thomas-jj.eurosky.social · 22/07/2026
En gros, pour résumer les bruits convergents, on aurait : - INSIS+INP > CNRS Physique & Ingénierie - INSMI+INS2I > CNRS Mathématique & Informatique - INEE+INSU (climat-océan) > CNRS Terre & Biosphère - IN2P3+INSU (espace) > CNRS des 2 infinis
474
Reposted by Nabil Hathout
Naomi Saphra @nsaphra.bsky.social · 20/07/2026
😭
Peter Henderson
@PeterHndrsn
Several people with no technical background told me they submitted to NeurIPS using agents for all coding—stating they didn't understand what the agent did, nor underlying code. 

Asked how they could trust the code/results, they said they assume reviewers will catch any errors.
49618
Reposted by Nabil Hathout
Danny Kingsley @dannykay68.bsky.social · 20/07/2026
EXCELLENT: The Netherlands new OA policy seems to be well thought through obc.copim.pub/when-scaling... Policy explicitly recommends redirection & reallocation of funds from existing ‘big deal’ budgets towards community-driven alternatives, incl university-led diamond OA & Netherlands Uni Presses
obc.copim.pub
When 'scaling small' becomes national policy
The Netherlands' new open access strategy echoes the principles the Open Book Collective was built on - 'scaling small' and community-led publishing
17333
Reposted by Nabil Hathout
Christophe Nahon @chnahon.bsky.social · 18/07/2026
Le Centre national de la recherche scientifique (CNRS) annonce l’accélération de sa sortie de certains logiciels extra-européens pour gagner en souveraineté numérique et réduire ses coûts IT. www.zdnet.fr/actualites/l...
zdnet.fr
Le CNRS sort de Zoom et Microsoft Exchange, et bientôt Sharepoint - ZDNET
Le Centre national de la recherche scientifique annonce l’accélération de sa sortie de certains logiciels extra-européens pour gagner en souveraineté numérique et réduire ses coûts IT.
033
Reposted by Nabil Hathout
Eugene Vinitsky 🍒 @eugenevinitsky.bsky.social · 17/07/2026
One thing I did not understand before starting as a faculty was just how important the NSF is. It's one of the few sources of nearly open-ended research funding that exists in the US. Everything else is significantly more directed
2693
Reposted by Nabil Hathout
The Data Therapist in the Blue Sky @datatherapist.bsky.social · 18/07/2026
Context; aclrollingreview.org/may26-cycle-...
aclrollingreview.org
An explanatory letter on the ARR May 2026 cycle
For the ARR May 2026 cycle, we had to make the difficult decision to increase the reciprocal service assignments for submitting authors. The reason for this decision was the substantial increase in su...
001
Reposted by Nabil Hathout
Hetan Shah @hetanshah.bsky.social · 14/07/2026
‘Thousands of university job cuts in humanities and social sciences are creating widespread cold spots for languages, classics and theology degrees, the British Academy has warned’ www.theguardian.com/education/20...
theguardian.com
Humanities among degrees being ‘extinguished’ by hard-up universities
Exclusive: Experts fear rise in education ‘cold spots’ and social immobility as 4,000 academic posts lost in one year
7181133
Reposted by Nabil Hathout
David Jurgens @davidjurgens.bsky.social · 13/07/2026
With my ACL PC abilities, I wanted to characterize who these folks are, whether they're actually "outsiders", whether their papers are much less likely to be accepted, is this growth just LLM slop? The whole analsis is here medium.com/@jurgens_245...
medium.com
Is the ACL Rolling Review actually broken?
ARR is broken! ARR is being overwhelmed with papers from outsiders! LLM-generated slop is ruining our peer review! There are not enough…
12811
Reposted by Nabil Hathout
langscipress.bsky.social @langscipress.bsky.social · 08/07/2026
Just published 'Voids in morphology: Exploring “uninflectedness” ' edited by Enrique L. Palancar & Sebastian Fedden #openaccess #eotms langsci-press.org/catalog/book...
book cover
092
Reposted by Nabil Hathout
Marzena Karpinska @markar.bsky.social · 06/07/2026
Only 17% of May #ARR authors are qualified for reviewing, and almost 40% of papers have no qualified people?!?!
2304
Reposted by Nabil Hathout
Linguistic Society of America @lingsocam.bsky.social · 29/06/2026
What happens when a linguistics program faces suspension amid enrollment & budget pressures? At Cleveland State University, faculty were facing just that situation. Learn how they did it! Join us on July 15! www.lsadc.org/ev_calendar_... #LSA #Webinar #Linguistics
032
Reposted by Nabil Hathout
Djamé 🟥 @zehavoc.bsky.social · 15/06/2026
Tout à fait, pour la petite histoire un mec a compilé des infos sur les recrutements d'Anthropic. C'est assez éclairant : ils ne recrutent principalement que des pros (donc payés très cher), des ingénieurs, pas tellement des docteurs. C'est assez logique: faire un modèle c'est la première étape
1. Anthropic a monté son équipe presque du jour au lendemain
Parmi les ingénieurs encore chez Anthropic, seulement 15 sont arrivés avant 2021.
La vraie grande expansion, c'est en 2025-2026. En 2025, l'équipe ingénierie a triplé en un an, avec 686 recrutements. En 2026, le rythme semble similaire : jusqu'en juin, ils en ont déjà embauché 455.
Aujourd'hui, la moitié de l'équipe a moins d'un an d'ancienneté ; 53 % sont arrivés au cours des 12 derniers mois. La médiane de temps en poste est de 10 mois.
En clair, c'est une équipe géante montée à toute vitesse en à peine 18 mois.

2. Ils recrutent presque exclusivement des ingénieurs seniors
C'est le point qui m'a le plus surpris, le plus contre-intuitif.
Avant d'arriver chez Anthropic, l'expérience professionnelle médiane de ces personnes est de 12,2 ans. Pour les 50 % du milieu, ça va de 8,8 à 16,5 ans.
Sur les 1680 personnes, seulement 50 ont moins de 3 ans d'expérience ; 44 % en ont plus de 13 ans. Les sortants d'école, quasiment zéro.
Donc, un nouveau typique chez Anthropic, c'est quelqu'un qui a déjà 12 ans de carrière, mais qui est dans l'entreprise depuis seulement 10 mois.

3. Ils privilégient l'infra, pas la « recherche » comme on pourrait le penser
40 % des profils mentionnent de l'infrastructure.
Backend, systèmes distribués, bases de données, sécurité : chacun autour de 20 %.
Tandis que le reinforcement learning, seulement 3,3 % des personnes.
En gros, un ingénieur typique chez Anthropic a passé la dernière décennie à monter des systèmes de production à grande échelle dans des hyperscalers ou des startups infra-intensives.

4. La principale source de talents n'est pas un labo IA, c'est Google
Tout le monde imagine qu'Anthropic recrute beaucoup chez OpenAI et DeepMind.
En réalité, son plus gros vivier, c'est Google, et de loin.
À part Google, ils affectionnent clairement les endroits réputés pour leur rigueur en ingénierie : Stripe, Databricks, Snowflake, Palantir, Airbnb
222
Reposted by Nabil Hathout
Djamé 🟥 @zehavoc.bsky.social · 15/06/2026
Non, on a absolument pas les données instruct et d’alignement qu’il faudrait. Le seul moyen pour s’en approcher serait de distiller massivement et sans pitié les modèles frontières. Ce qu’auraient fait les chinois avec Deepseek par exemple.
332
Reposted by Nabil Hathout
langscipress.bsky.social @langscipress.bsky.social · 15/06/2026
Language Science Press has published 300+ books in the last 12 years. Every 3 years, our funding has to be renewed. We need 100 institutions giving us €2000/year to continue our operations in 2027. Please visit langsci-press.org/knowledgeunl... for more information and talk to your librarian.
langsci-press.org
Funding | Language Science Press
089
Reposted by Nabil Hathout
Michael Okun @michael-okun.bsky.social · 10/06/2026
I’ve officially resigned as Associate Editor for Frontiers in Systems Neuroscience. It used to be a reputable journal, but became a case study in how forced automation destroys academic integrity. 👇
5435871430
Reposted by Nabil Hathout
Laurent Besacier @lbesacier.bsky.social · 03/06/2026
As Claude and similar tools become serious research accelerators, a new divide is emerging: industrial labs can afford enterprise AI plans for their scientists; many academic labs cannot. The research gap isn't just about compute anymore. Not everyone gets the shot.
011
Reposted by Nabil Hathout
Arnaud Raoux @physicochimix.bsky.social · 31/05/2026
www.lesnumeriques.com/informatique... "Le logiciel libre l’emporte : Polytechnique dit non à Microsoft"
lesnumeriques.com
Le logiciel libre l’emporte : Polytechnique dit non à Microsoft et déclenche une vague de rébellion dans les universités
C'est un coup de théâtre qui secoue le monde de l'enseignement supérieur et de la tech française. Après des mois d'une guerre de tranchées juridique, la direction de l'École polytechnique a officielle...
512154
Reposted by Nabil Hathout
Alexis Verger 🧬🧫🧪 @alexis-verger.cpesr.fr · 22/05/2026
😬 #CNRS #ESR www.aefinfo.fr/depeche/7511...
aefinfo.fr
CNRS : Emmanuel Macron propose de nommer Thierry Dauxois comme PDG
035
Reposted by Nabil Hathout
David Monniaux @monniauxd.bsky.social · 20/05/2026
Food for thought harpers.org/2015/12/who-...
harpers.org
Who Goes Nazi?, by Dorothy Thompson
Donald Trump calls for a ban on Muslims entering the United States; Dorothy Thompson wonders who is most susceptible to Nazism.
253
Reposted by Nabil Hathout
CoNLL 2026 @conll-conf.bsky.social · 18/05/2026
Hallucinated references are the talk of the town now. How did we go about paper submissions with "hallucinated" references? At CoNLL, we viewed incorrect references as a violation of fundamental scholarly standards, and papers with such references were not sent for review.
1113