It is time to abandon student evaluations of teaching Comment by Gordon Hodson (@gordonhodsonphd.bsky.social) Web: go.nature.com/4jfAzXo PDF: rdcu.be/ef9y5

In this commentary, I argue that student-based teaching evaluations are problematic bc ✳️ systematic biases (e.g., racism, sexism) ✳️ poor construct validity ✳️ undermine standards and learning We should evaluate teaching as seriously as we do research. Or don't do teaching evaluations.

Fortement d’accord avec ce point de vue. Je travaille dans le domaine, mais j’ai horreur des extrêmes absurdes auxquels songent certaines personnes.

Ready to vote. Prêt-e-s pour voter.

La simple question de savoir pourquoi ouvre des perspectives terrifiantes.

Are we going to witness a real bubble burst? Or a slow, boring deflation instead?

Monday again...

Claus Wilke on Alphafold and the problem of protein folding in 2025

AlphaFold is great, but contrary to public opinion it has not completely solved the protein folding problem. Much work remains to be done. clauswilke.substack.com/p/no-alphafo...

https://www.theguardian.com/commentisfree/2025/jul/08/governments-stand-up-to-israel-colombia.

"We are perilously close to the point of no return" - climate scientist on the future of the Amazon rainforest. The world's leaders have seemingly unanimously agreed that keeping a liveable biosphere is less important than ensuring the richest people get a little richer.

Springer-Nature launched a series of "Discover" journals that closely mimic MDPI titles -- sharing *identical* journal names, and likely similar business model. What is going on, and why researcher will - as always - fall for it? A 🧵 the-strain-on-scientific-publishing.github.io/website/post...

"a fascinating case study on the limits of AI in biology and the harms of current publishing incentives" rachel.fast.ai/posts/2025-0...

#drosophila @flybase.bsky.social request emergency funding: "As it stands, by the end of July, 2025, there will be no future updates to FlyBase, and in the worst case scenario access to the website will also be lost" => please donate! www.philanthropy.cam.ac.uk/give-to-camb...

Should LLMs be used to review papers? AAAI is piloting LLM-generated reviews this year. I wrote a blog post arguing that using LLMs as reviewers can have bad downstream consequences for science by centralizing judgments about what constitutes good research. bryanwilder.github.io/files/llmrev...

🧪🔬 On the downsides of funding megaprojects and why we need to reconsider power relationships when funding science. I guess if @snsf.ch #SNSF has any hard data on the comparative impact that #NCCR projects have on the careers of researchers at different stages in their career... #AcademicSky

This is a fantastic oral history of the last 10 years of NLP and AI. www.quantamagazine.org/when-chatgpt...

Lancement du projet interreg « GESICA »

Very informative chart from Washington Post

La #Suisse veut à nouveau renvoyer des demandeurs d’asile déboutés en #Afghanistan – du moins s’ils ne sont pas « vulnérables ». Pourtant, ce pays demeure dangereux, indépendamment du genre ou de l’âge. La Suisse devrait accorder sa protection, et non pas expulser ! www.admin.ch/gov/fr/accue...

Le Conseil des États a rejeté aujourd’hui une motion visant à établir un cadre législatif pour encadrer le secteur des matières premières. 🟥 Sans loi spécifique, la Suisse continuera à tirer profit de sa place de négoce, principalement aux dépens des populations dans les pays producteurs. 👇

🔥 Unveiling the Future of Genomics with Genome Language Models (gLMs)! 🔥 Our comprehensive review, "Transformers and genome language models," is finally published in Nature Machine Intelligence! ​ Link: nature.com/articles/s42...

Triste retour à l'horreur. lecourrier.ch/2025/03/18/u...

I've recently put together a "Fairness FAQ": tinyurl.com/fairness-faq. If you work in non-fairness ML and you've heard about fairness, perhaps you've wondered things like what the best definitions of fairness are, and whether we can train algorithms that optimize for it.

A #bioinfo @xkcd.com comic. Nice.

Interesting proposal (cheap technical solution) to improve the whole EV Charging chain. Thus improving EV adoption. spectrum.ieee.org/ev-charging-...

“Congress shall make no law respecting an establishment of religion, or prohibiting the free exercise thereof; or abridging the freedom of speech, or of the press; or the right of the people peaceably to assemble, and to petition the Government for a redress of grievances”

🤦‍♂️Sad but realistic. Is NLP a doomed research domain?

Nice morning pattern

Major cuts planned in SRF science journalism in Switzerland - the public service media. Scientists today launched a petition ‘Facts Instead of Misinformation: For Strong Science Journalism!’ Become one of over 1,800 signatories www.openpetition.eu/ch/petition/...

Moi aussi ✍️

Can LLMs replace human teaching assistants in grading homework assignments? This semester, we ran a study in our Bioinformatics course with 100+ enrolled master's students, whose assignments were graded by 6 LLMs. 1/

To quote @denisduboule.bsky.social , IMHO one of the best definition of curiosity-driven basic research & serendipity "going off to look for mushrooms and coming back with blueberries. If you are payed to find the mushrooms, you won't come back with the blueberries. You will come back with nothing.

I really think many if not most of the LLM fail cases are people trying to use it as a search engine. It is the opposite of a search engine.

I honestly believe all LLM results should come with a disclaimer reminding people that the thing doesn’t (and absolutely cannot) know any facts or do any reasoning; it is simply designed to *sound* like it knows facts and frequently sweeps some up accidentally in the process.

No more comments. lecourrier.ch/2024/12/04/c...

Seems this paper is making some waves as it illustrates what effects may AI have on scientific productivity (hint: it depends)

Le dessin d’il y a 2 ans remis à jour pour Vert. Un dessin que vous ne retrouverez pas dans le chapitre « écologie » du futur recueil de dessins de l’année 2024 mais pour vous consoler vous pouvez le télécharger pour l’afficher depuis le site de Vert !

“Again and again, our major institutions, from the media to the judiciary, have amplified Trump’s presence; again and again, we have failed to name the consequences. Fascism can be defeated, but not when we are on its side.”

Nous sommes désormais plus de 10 millions sur Bluesky, et j’étais n°2 558 405 !

Dritchino festival

We would need better data and models, yes. But, in the meantime, also act as if worst case scenarios were real.

"Climate models can’t explain 2023’s huge heat anomaly — we could be in uncharted territory"

www.linkedin.com/posts/activi... IT FINALLY HAPPENED! THE AI ACT HAS JUST PASSED! By Barry Scannell.

BioLLMBench: A Comprehensive Benchmarking of Large Language Models in Bioinformatics https://www.biorxiv.org/content/10.1101/2023.12.19.572483v1

Good morning #episky #publichealth #medsky #pedsky #academicsky #womeninSTEM 🧪 Credit: @xkcd.com