Daniel Saraga @danielsaraga.eurosky.social · 14/10/2026but this holds only if the above list is really globally shared... one aspect is the globalization of issues. Even if an issue is local (eg population growth in CH/LU or decrease of public service in UK), people might hear about it and absorb it as their problem, making it a virtual local issue 100
Daniel Saraga @danielsaraga.eurosky.social · 14/10/2026My bet is a mix of the perceived... –clash of civilisations –immigration –culture wars –climate catatrophe –sense of "the future won't be better" –loss of communal feeling –increased individualism, fueled by digitization –loss of agency 100
Daniel Saraga @danielsaraga.eurosky.social · 14/10/2026Mh, my feeling is that local explanations for the rise of the extreme right vote can only go that far. The rise is widespread in so many western countries that so I rather bet on macro, global explanations. 100
Daniel Saraga @danielsaraga.eurosky.social · 13/10/2026In Switzerland the public service is very good but the party with most votes has been for years the SVP, which extends comfortably into far right territory. A counter example? I have no idea, it's just a cheap thought :-0 100
Daniel Saraga @danielsaraga.eurosky.social · 08/10/2026In effect, this disrupt our capacity to type in what we want in a machine. The machine demands from us that we describe what we want before interpreting it. This means the #AI system is in charge, not the human any more. It's weird. And, on a more symbolic level, completely unsettling. 000
Daniel Saraga @danielsaraga.eurosky.social · 08/10/2026I am a bit puzzled by the change operated by #OpenAI when they transformed #GPTs into #Plugins. Now, the user cannot edit directly the instructions of a plugin anymore. Instead, it must interact with ChatGPT: asking it to list the instructions and then asking it to modify it. 1/2 100
Reposted by Daniel SaragaChris Hayes @chrislhayes.bsky.social · 07/10/2026No AI for kids. Full stop. 2245281913
Daniel Saraga @danielsaraga.eurosky.social · 06/10/2026je crois que le pire est au niveau meta : il repousse sans cesse les limites de qui devrait normalement être une ligne rouge. c'est la force de l'absurde. en France on se demande si un politicien va ou non souffrir de ses mensonges. probablement pas, si on continue à suivre les US. Anything goes 010
Daniel Saraga @danielsaraga.eurosky.social · 05/10/2026Je t'encourage! Tu as pas mal de chances de pouvoir commencer sans avoir fait la formation. Elle n'est pas strictement obligatoire, mais assure de meilleures conditions salariales. Selon les cantons, elle peut se faire à temps partiel. 010
Daniel Saraga @danielsaraga.eurosky.social · 25/09/2026on the mesoscale: mimilette cheese comes with mites included 010
Daniel Saraga @danielsaraga.eurosky.social · 24/09/2026yep. j'ai le même sentiment par rapport l'IA. Les gens semblent paralysés devant la problématique de l'IA. Quant OpenAI hacke les données médicales australiennes, l'Etat ne porte pas plainte comme il le ferait contre n'importe quel groupe de hackers normal. Arg là je mélange tout, désolé! 000
Daniel Saraga @danielsaraga.eurosky.social · 24/09/2026damn, tout ça va raviver l'intérêt de versionner les documents littéraires avec une blockchain, cette polluante technologie en quête d'un problème à résoudre. 010
Daniel Saraga @danielsaraga.eurosky.social · 24/09/2026*j'imagine que le versionage des documents qui pourrait démontrer un rythme "humain" dans la succession des différentes version d'un document pourrait aider à dissiper le doute enfin jusqu'à ce que les agents apprennent à générer du texte avec un rythme humain.. 100
Daniel Saraga @danielsaraga.eurosky.social · 24/09/2026sous réserve en effet. Un accusation d'utilisation de l'IA n'est en général étayée que du résultat très discutable d'une analyse faite par une deuxième IA. * Un plagiat, lui, est visible de manière objective. 100
Daniel Saraga @danielsaraga.eurosky.social · 28/08/2026Amazing Franck, what an idea! Wish you all the best! 010
Daniel Saraga @danielsaraga.eurosky.social · 23/07/2026pas mal cet article: medium.com/@rantnrave31...medium.comFritz Zwicky: The Astronomer Who Was Right About Everything (And Nobody Listened)A Retroactive Nobel Prize Nomination for Being Annoying and Correct 000
Reposted by Daniel SaragaJean Abbiateci @jeanabbiateci.bsky.social · 22/06/2026Sur @heidi-news.bsky.social, j'ai raconté la genèse de mon enquête sur Etienne Klein, depuis l'enquête sur sa thèse plagiée avec @lorisguemart.bsky.social pour @arretsurimages.net ➡️ www.heidi.news/explorations... . 24216
Daniel Saraga @danielsaraga.eurosky.social · 17/06/2026Merci! Une question: est-ce que ce test a tenu compte des différences dans la loi (et de la sévérité des peines prononcées) en France et Suisse? Je ne l'ai pas vu dans l'encadré Méthodologie. 000
Daniel Saraga @danielsaraga.eurosky.social · 16/06/2026Extremely impressive. The system guesses the phonemes that the person has been trying to vocalize and auto-correct and -complete them into words and sentences. 000
Daniel Saraga @danielsaraga.eurosky.social · 12/06/2026I cannot see of any possible safety risks here, but it's vibe-coding made by an incompetent coder, so I hope I will not contribute to an even unsafer internet. Use it at your own risk. 000
Daniel Saraga @danielsaraga.eurosky.social · 12/06/2026➡️ To install it: create a new bookmark in your browser > edit its URL > change it to the script. ✅ To use it: search on sbb.ch (EN, FR or DE) > click on the desired connection (you must do it) > click on the bookmark > paste the text where you want #VibeCodingsbb.chDas Online-Portal der SBB für Fahrplan, Zug und ÖVAlles über Halbtax, GA, Billette, Ausflüge, Städtereisen und Event-Tickets sowie Infos zu Bahnhöfen, Störungen & Fahrplan. 110
Daniel Saraga @danielsaraga.eurosky.social · 12/06/2026I wanted a tool to copy-paste trains connections from the sbb.ch website to my calendar in a simple text-format I have asked claude and ChaGPT to cook up a solution. It's a bookmarklet. The code to put in the bookmarklet's url is in this document: docs.google.com/document/d/1... 110
Reposted by Daniel SaragaFranck Marchis @allplanets.bsky.social · 10/06/2026Tomorrow, I’m hosting a live Reddit AMA! Ask me anything about astronomy, SETI, SkyMapper, SkySphere, citizen science, UAP data, or our goal of building a global network that continuously observes the sky. 📅 June 11 ⏰ 11 AM PT / 2 PM ET www.reddit.com/r/IAmA/comme...reddit.comFrom the IAmA community on RedditExplore this post and more from the IAmA community 0133
Reposted by Daniel SaragaFranck Marchis @allplanets.bsky.social · 09/06/2026We are very close to reach our goal on @kickstarter.com with our SkySphere, an intelligent all-sky camera designed to detect natural phenomena in the sky, as well as human-made ones or even extraterrestrial ones... Join us! #SETI #UAP @setiinstitute.bsky.social 2358
Daniel Saraga @danielsaraga.eurosky.social · 04/06/2026The irony: the leaked/transmitted protocols were about implementing the EU's Digital Service Act 000
Daniel Saraga @danielsaraga.eurosky.social · 02/06/2026Generating anti-AI messages with AI. We humans either have no shame, or delight in irony. 000
Daniel Saraga @danielsaraga.eurosky.social · 27/05/2026argh, sorry to hear. pour rejoindre les gens qui donnent des conseils sans avoir été solicités: un bassine d'eau froide (et des glaçons pourquoi pas) sous le bureau pour y plonger les pieds, ça marche assez bien. l'air con (qui porte peut-être bien son nom :-0) écolo 000
Daniel Saraga @danielsaraga.eurosky.social · 18/05/2026Thanks, interesting. A quick question: was there a control for other (e.g. socioeconomic) variables that could influence the outcomes? Many thanks. 100
Reposted by Daniel SaragaSwiss National Science Foundation (SNSF) @snsf.ch · 27/04/2026#SwissScienceImage 🦗 📸💎 A cricket's leap, a ballet of phytoplankton and the beauty of vitamin B crystals: Discover the 20 winning photos and videos from the SNSF Scientific Image Competition 2026. For more info: www.snf.ch/en/1uKR6zjXHw… 043
Daniel Saraga @danielsaraga.eurosky.social · 24/04/2026More science communication workshops: ➡️ lnkd.in/e5QaTAkS 5/5 000
Daniel Saraga @danielsaraga.eurosky.social · 24/04/20263️⃣ Use every opportunity to improve your communication skills: – open days – your department's newsletter – social media – grassroots initiative 4/5 100
Daniel Saraga @danielsaraga.eurosky.social · 24/04/20262️⃣ Distinguish: – what you know from your worn work – what is established knowledge from your field – your views as an expert – your views as a citizen 3/5 100
Daniel Saraga @danielsaraga.eurosky.social · 24/04/20261️⃣ Always think about: –Why am I communicating? –Who is this for? –What do they need? –Which format and channel fit best? –What do I know? –How should I communicate it? 2/5 100
Daniel Saraga @danielsaraga.eurosky.social · 24/04/2026#ScienceCommunication #101 ! > When to talk about your research to a broader audience? > How to write concise posts about your results? > What to do upon a media request? I have worked on these topics with PhD students at @unibas.ch during a 1-day workshop on science communication. 1/5 111
Daniel Saraga @danielsaraga.eurosky.social · 17/04/2026and also for researchers. A 12-15% success rate is hard but might be still OK. Below 10% one can wonder if it's worth it. The Swiss SNSF's lottery for borderline applications in certain schemes (neither outstanding, nor clearly insufficient) is a weird, but interesting idea. 000
Daniel Saraga @danielsaraga.eurosky.social · 17/04/2026mh... One has to understand the perspective of ERC: too many applications can be a problem. The ERC budget seems* to have increased more or less on par with the size of the pool of researchers* *(courtesy of a well know LLMS, numbers not checked). 100
Daniel Saraga @danielsaraga.eurosky.social · 12/04/2026Your tolerance for showers that feel like a light drizzle. 020
Daniel Saraga @danielsaraga.eurosky.social · 26/03/2026Tous ces plagiats, c'est un peu comme des paper mills, mais faits en artisan. 110
Daniel Saraga @danielsaraga.eurosky.social · 26/03/2026comme quoi (bis repetita) le monde académique semble mieux discipliné pour ce genre de problème, avec le système des rétractions... 020
Daniel Saraga @danielsaraga.eurosky.social · 26/03/2026mh. on imagine une banque qui emploie une personne accusée de fraude et qui répond "c'était avec la direction précédente"? Oui peut-être, mais donner l'impression de ne pas se sentir concerné, ce n'est pas terrible pour la confiance des clientes et auditrices, non? 1101
Daniel Saraga @danielsaraga.eurosky.social · 26/03/2026Comme quoi, le milieu scientifique n'est pas toujours le plus incapable à gérer les cas problématiques. Ceci dit, les autres silences radio ne sont guère étonnants, en général ce genre d'enquête se fait de manière discrète. Si elle se fait. 130
Daniel Saraga @danielsaraga.eurosky.social · 25/03/2026Fascinating talk by Cory Doctorow about a small window of opportunity to sever our total (delusional?) dependence on US technology. media.ccc.de/v/39c3-a-pos... @doctorow.pluralistic.net @ccc.demedia.ccc.deA post-American, enshittification-resistant internetTrump has staged an unscheduled, midair rapid disassembly of the global system of trade. Ironically, it is this system that prevented all... 2213
Reposted by Daniel SaragaCaroline Orr Bueno @rvawonk.bsky.social · 25/03/2026Former WH Chief Strategist Steve Bannon suggested on his podcast this week that deploying ICE agents to airports is a “test case” for sending them to polling places on Election Day. History tells us that it’d be wise to take this seriously. 54023
Daniel Saraga @danielsaraga.eurosky.social · 05/03/2026damn, the truth about truths is behind a paywall 120
Daniel Saraga @danielsaraga.eurosky.social · 04/03/2026Thanks for the hint. Indeed, and there is no flag about this in the online version today (a month later) www.nature.com/articles/s41... This has been flagged on PubPeer pubpeer.com/publications... Sci Rep has work to do, and fast.nature.com 100