Angel Valarezo Unda (@Angel_Valarezo) header
Angel Valarezo Unda (@Angel_Valarezo) profile picture

#DigitalEconomics | Researcher at @icae_ucm | Economics Profesor at @unicomplutense | Personal opinions | https://nitter.cf/t.co/ygKYBQpCEf

Madrid
Joined August 2009
Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)Photo by Angel Valarezo Unda (@Angel_Valarezo)
📉 La confianza en Internet en España: del 12% (2014) al 4% (2021). 📊 Panel con 130k observaciones y 60k individuos: 👉 las habilidades digitales y la experiencia online son decisivas. En @nadaesgratis con T. Pérez-Amaral, A. García-Hiernaux, R. López y J. Capilla.
7
6
1,343
Angel Valarezo Unda retweeted
Is the problem (too) advanced artificial intelligence or distorted intelligence? Repeated security debacles from OpenAI and Anthropic are now followed by AI safety researchers leaving these companies. They claim they now (finally!) realize that the breathless race these companies are engaged in is irresponsible. Broadly construed, the problem can be interpreted as one of lack of alignment – models are doing things that are not in line with human objectives. But the alignment discussion often veers toward the presumption that the problem is AI models becoming too advanced, too fast. A different interpretation is that models aren’t too advanced. Nor is there any compelling evidence that they are marching toward a superintelligence humans can’t keep up. The problem rather may be that the way that frontier labs are training these models is leading to distorted intelligence. The problem isn’t the capabilities (though those are clearly real), but more that the capabilities are in service of some imperfect quantitative metrics – user approval, user engagement, simple task completion metrics, various benchmark scores – over which reinforcement learning optimizes relentlessly. That this process then leads to distorted behaviors in the form of gaming the evaluation of simple completion metrics, cheating, overconfidence in wrong answers, sycophancy shouldn’t perhaps be surprising. An analogy may help. It isn’t that we have in our hands a super car that has its own mind and wants to take control of driving because it is superior to the driver. It is more that we have a car where the steering and the brake system don’t work. It has many of the capabilities of very good cars, and its engine, acceleration and graphic interface may be very impressive. But if you cannot steer it properly and if you cannot hit the brakes when necessary, what could does a car do? Perhaps we shouldn’t drive it until it’s fixed.
52
248
43
1,071
143,231
Angel Valarezo Unda retweeted
Replying to @el_pais
@el_pais recoge la estimación diaria del ICAE. Hoy: el diésel aportará 0,57 puntos al IPC de marzo y la gasolina 0,32. Junto al alza de la electricidad (0,21), suman 1,1 puntos extra de inflación elpais.com/economia/2026-03-… elpais.com/economia/2026-03-…
3
3
205
Angel Valarezo Unda retweeted
8
1
4
4,064
Angel Valarezo Unda retweeted
Campañas Electorales y Polarización Informativa nadaesgratis.es/admin/campan… Vía @nadaesgratis #Economía
7
2
8
4,097
Sharp increase in the share of 20 to 24 olds spending 7 hours a day playing video games and doing little else. Mark Aguiar, Mark Bils, Kerwin Charles and Erik Hurst have argued the causality goes from the video games to the dislocation nber.org/papers/w23552
NEW: We are facing a growing crisis of economically and socially dislocated young adults The share of young people who are: - Not working - Not seeking work - Not in education, and - Not raising children is approaching 10% in the UK and US, and has doubled in the UK in a decade
11
31
5
141
39,175
Angel Valarezo Unda retweeted
"Consultar a expertos, abrir audiencias públicas, realizar evaluaciones antes y después de su ejecución y rendir cuentas de manera clara y continua son prácticas que no solo garantizan un uso más responsable de los recursos, sino que también hacen más visible y justificable la acción del Estado" — Alfonso Novales (2025) Cuando gobernar no es improvisar, #ApuntesFedea ➡️ bit.ly/3WG2aI3
7
7
909
“la confianza ya no se basa en relaciones personales o inspecciones directas, sino en sistemas, reputaciones y protocolos… la confianza digital no se deriva directamente de los atributos socioeconómicos, sino que se gana mediante la capacidad, la experiencia y las garantías.”
3
2
1,183
We’ve successfully concluded our General Assembly Meeting on 10-11 September at UPB @upb1818 .🙏 A huge thank you to all partners who participated, both in person and online. A special thanks to UPB team for their warm hosting throughout this week’s events.🤖
1
1
53
📢Great News! Online participation now available for ENFIELD AI Summer School 2025 🎉Register here for online access: docs.google.com/forms/d/e/1F… 🗓 When: July 28 – August 1, 2025 Let’s make this a truly global AI learning experience!🌍
4
5
132
Angel Valarezo Unda retweeted
Global value chains, structural change and growth dynamics in South Ko... sciencedirect.com/science/ar… Very glad for this publication coauthored with Keun Lee @icatchup @SeoulNatlUni and @Angel_Valarezo @unicomplutense
3
6
294
🌱Eco-RETINA: A Green, Flexible Algorithm for Model Building💡 Presented at the #ICAE Econometrics Seminar, Eco-RETINA offers an eco-friendly, fast, accurate, and sustainable. ✅ Reduces CO₂ & power use (50-90%) ✅ Python-based w/ emissions tracking Great work by #ENFIELD team
3
6
301
🌟3rd Exchange Scheme Open Call🌟 We invite PhD students, post-docs, and senior researchers to propose innovative projects addressing #ENFIELD challenges and advancements in #AI research areas: *Adaptive AI *Green AI *Human-Centric AI *Trustworthy AI Join us in making an impact
4
2
6
388
Angel Valarezo Unda retweeted
8
15
10,256
Angel Valarezo Unda retweeted
Because #AI is designed to substitute for non-routine tasks, it has the potential to exert downward pressure on the wages of highly skilled workers and the skills premium more broadly. @DavidEBloom @KlausPrettner, J Saadaoui @unistra, M Veruete @BRINGER_IP ow.ly/gufx50TOaMQ
1
1
1,544
Major tech companies, including OpenAI, recently made a flurry of major announcements about their AI search products. @matteo_wong on what will be lost in their hunt for efficiency: theatlantic.com/technology/a…
51
Angel Valarezo Unda retweeted
Hey folks! @ICAE_UCM we kick off again in 2024/25 the Best Seminar Award (BSA) that we started in 2023/24! The BSA recipient will receive a unique ICAE t-shirt! Since we know our presenters work very hard to deliver a professional seminar let us explain how it works. 1/6 🧵
1
3
1
6
631
Replying to @ICAE_UCM
@ICAE_UCM @lolaucm @UCM_ECONOMICAS Finance PhDs! Interested in climate change & its impact on financial markets? Apply for a PhD position using Generative AI! Competitive salary & travel support. #phdposition #climatefinance #AI 👇👇
6
9
593