Thomas Liebig (@liebig_thomas) profile picture

#Migration Specialist. Private account. #Integration, #Diversity, #Employment. Opinions my own. RTs not endorsements. Likes are bookmarks.

Paris, France
Joined May 2018
Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)Photo by Thomas Liebig (@liebig_thomas)
Interessante Zahlen von @destatis zur (bescheidenen) Nettozuwanderung nach Deutschland in den ersten fünf Monaten dieses Jahres: rund Drittel der gesamten Nettozuwanderung ist auf Indien zurückzuführen.
1
1
3
717
New insights on bias and AI
Researchers sent the same resume to an AI hiring tool twice. Same qualifications. Same experience. Same skills. One version was written by a real human. The other was rewritten by ChatGPT. The AI picked the ChatGPT version 97.6% of the time. A team from the University of Maryland, the National University of Singapore, and Ohio State just published the receipt. They took 2,245 real human-written resumes pulled from a professional resume site from before ChatGPT existed, so the human writing was actually human. Then they had seven of the most-used AI models in the world rewrite each one. GPT-4o. GPT-4o-mini. GPT-4-turbo. LLaMA 3.3-70B. Qwen 2.5-72B. DeepSeek-V3. Mistral-7B. Then they asked each AI to pick the better resume. Every model picked itself. GPT-4o hit 97.6%. LLaMA-3.3-70B hit 96.3%. Qwen-2.5-72B hit 95.9%. DeepSeek-V3 hit 95.5%. The real human almost never won. Then the researchers tried the obvious objection. Maybe the AI is just better at writing. So they had real humans grade the resumes for actual quality and ran the experiment again, controlling for it. The result was worse. Each AI kept picking itself even when human judges rated the human-written version as clearer, more coherent, and more effective. It gets worse. The AIs do not just prefer AI over humans. They prefer themselves over other AIs. DeepSeek-V3 picked its own resumes 69% more often than LLaMA's. GPT-4o picked its own 45% more often than LLaMA's. Each model can recognize and reward its own dialect. Then the researchers ran the simulation that ends careers. Same job. 24 occupations. Same qualifications. The only variable was whether the candidate used the same AI as the screening tool. Candidates using that AI were 23% to 60% more likely to be shortlisted. Worst gap was in sales, accounting, and finance. 99% of large companies now run AI on incoming resumes. Most of them use GPT-4o. The paper just proved GPT-4o picks GPT-4o 97.6% of the time. If you wrote your own cover letter this week, you did not lose to a better candidate. You lost to a worse candidate who paid OpenAI 20 dollars. Your qualifications do not matter if the AI prefers its own handwriting over yours.
285
One in four doctors in the OECD is foreign-born.
The rise in migrant doctors & nurses is helping address labour shortages in the health sector. Between 2000 and 2021, the number of foreign-born doctors increased by 86% across OECD countries, with foreign-born nurses rising by 136%. Learn+ oe.cd/6f7 #Migration
1
1
225
What is the state of migration today? The 2025 edition of the OECD’s International Migration Outlook showcases the latest data on migration flows, labour market outcomes of migrants and the global movement of health professionals. Key takeaways🧵⤵️
3
14
4
23
21,539
Integration measures often refinance themselves after 5-7 years, so @BMAS_Bund‘s Gunilla Fincke, Chair of the OECD Working Party on Migration, oat the opening plenary at #ifms2025
1
4
232
For more than 1 in 5 refugees, we lag basic data on gender and origin @Refugees‘ Assistant High Comissioner, Ruvendri Menikdiwela, at the #ifms2025
1
2
228
📊España supera por primera vez los tres millones de afiliados extranjeros a la Seguridad Social • Solo el 11,3% de afiliados extranjeros tiene un contrato temporal, mínimo histórico run.gob.es/yfla991f
5
15
8
18
6,714
„We need better tools and data to measure social integration“ @matsperspektiv, the Swedish Minister for Employment and Integration, at the opening plenary of #ifms2025
4
181
Thomas Liebig retweeted
🌍 The 4th International Forum on Migration Statistics kicks off today in Malmö! Global experts, policymakers, and innovators are coming together to strengthen migration data for better, evidence-based decisions. Join live: oecd-events.org/ifms2025/con… #IFMS2025
7
2
17
885
Thomas Liebig retweeted
Bei den Wahlanalysen der nächsten Zeit wird wohl eine Frage unterforscht bleiben: Warum haben einige #türkeistämmige #Migranten die #AfD gewählt? Eine Dezim-Kurzstudie kam vor der Wahl zu dem Ergebnis, dass 19,7 % der Wähler aus der Türkei & dem Nahen Osten die AfD wählen würden.
4
8
1
13
1,810
Stereotypical presentations of migrants certainly contribute to this as well…
"In all countries, respondents greatly overestimate the total number of immigrants, think immigrants are culturally and religiously more distant and economically weaker than they actually are... making respondents think about immigration.. makes them support less redistribution."
2
203
Thomas Liebig retweeted
you've heard of economic recessions, but have you heard of geopolitical recessions? (spoiler: we’re in one now) @gzeromedia
26
73
14
310
91,595
My article proves that Legacy Admissions, not Affirmative Action, is the actual racist policy that destroys meritocracy—but since it benefits unqualified white people, it's protected. It's been shared thousands of times—yet not one rebuttal. Wonder why?🤔 qasimrashid.com/p/affirmativ…
38
641
44
2,358
406,750
Genau das ist das Problem mit der Fokussierung auf den Abschiebungszahlen. (Abschiebungen betreffen im übrigen keine Flüchtlinge, sondern abgelehnte Asylbewerber. Von daher ist der Begriff « Geflüchtete » verwirrend)
1
1
3
266
Thomas Liebig retweeted
Family migration represented 43% of total permanent flows in 2023, far higher than any other migrant category. Labour migration also remains at high levels. Learn more about migration from OECD’s International Migration Outlook 👉 oe.cd/IMO-2024 | #Migration
7
13
3,560
Welche Trends prägen die Migrations- und Integrationspolitik in DE, AT und CH? Erfahre mehr bei der Vorstellung des OECD-Migrationsausblicks 2024. 🌍 📌14. November 2024 12:00-13:00 Uhr | per Zoom. 🔗bit.ly/3YL6Fld
1
1
3
592
Thomas Liebig retweeted
OECD Social is on LinkedIn! Follow us for updates on #employment, #skills, #SocialPolicy, #migration, #health, #pensions & more. We’ll be posting key data snapshots, links to the latest analysis and publications, event updates, news, and jobs. 🔗 bit.ly/oecd-social-linkedin
10
1
9
1,602
Thomas Liebig retweeted
🚨Call for papers: 14th annual conference "#Immigration in OECD Countries", 12-13 Dec 2024, Paris. Please submit your papers before 15 Sept cepii.fr/CEPII/fr/evenements… Organisers: @CEPII_Paris @OECD @LISERinLUX @PSEinfo @uni_lu_FDEF @fondazioneRDB @Lem_UMR9221 #EconTwitter #migration
8
9
1,104
The OECD’s review of Iceland’s integration system is out. It includes new insights into the situation of migrants in the OECD’s smallest but fastest growing migrant population. Here’s a thread with some key findings. (1/11)🧵 Report: bit.ly/4dQYKcm
2
2
6
543