Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 26/02/2025"Artificial Intelligence and the Labor Market" Nice analysis of LinkedIn data to estimate the impact of AI on jobs between 2010 and 2023. It illustrates the complexity of AI economic impacts: what exposure taketh away, complementarities and productivity giveth. www.menakahampole.com 153
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 26/02/2025"Artificial Intelligence and the Labor Market" Nice analysis of LinkedIn data to estimate the impact of AI on jobs between 2010 and 2023. It illustrates the complexity of AI economic impacts: what exposure taketh away, complementarities and productivity giveth. www.menakahampole.com 153
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 14/02/2025Metascience and AI postdoctoral fellowship Cool @sloanfoundation.bsky.social opportunity for anyone interested in studying AI x science adoption and impact. It is also very aligned with the evidence and experimentation recommendation in our recent essay! sloan.org/programs/dig... 053
Reposted by Juan Mateos-GarciaDan Larremore @danlarremore.bsky.social · 13/02/2025Intl Conf on the Science of Science & Innovation abstracts open! 👇👇👇 042
Reposted by Juan Mateos-GarciaMaria del Rio-Chanona @maria-drc.bsky.social · 29/01/2025📢 𝗡𝗲𝘄 𝗽𝗮𝗽𝗲𝗿 alert! We document how 𝗚𝗲𝗻𝗲𝗿𝗮𝘁𝗶𝘃𝗲 𝗔𝗜 𝗶𝘀 𝗿𝗲𝘀𝗵𝗮𝗽𝗶𝗻𝗴 𝗱𝗲𝗺𝗮𝗻𝗱 𝗳𝗼𝗿 𝗳𝗿𝗲𝗲𝗹𝗮𝗻𝗰𝗲𝗿 𝘀𝗸𝗶𝗹𝗹𝘀. 💼 💻 📊 We analysed 𝟯𝗠+ 𝗷𝗼𝗯 𝗽𝗼𝘀𝘁𝘀, used embeddings to cluster them into 𝟭𝟬𝟬+ 𝗳𝗶𝗻𝗲-𝗴𝗿𝗮𝗶𝗻𝗲𝗱 𝘀𝗸𝗶𝗹𝗹𝘀, and LLMs to classify them as s𝘂𝗯𝘀𝘁𝗶𝘁𝘂𝘁𝗮𝗯𝗹𝗲, 𝗰𝗼𝗺𝗽𝗹𝗲𝗺𝗲𝗻𝘁𝗮𝗿𝘆, or 𝘂𝗻𝗮𝗳𝗳𝗲𝗰𝘁𝗲𝗱. 22714
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 30/01/2025"Antivenom Renaissance" Nice illustration of how AI can accelerate science by helping find needles in different types of haystacks. 1. AI is being used to design proteins that neutralise venoms. The haystack is the space of all possible protein designs. blog.asimov.com/p/antivenom-... 132
Reposted by Juan Mateos-GarciaDaniel Rock @danielrock.bsky.social · 29/01/2025Hi Everyone! We're hosting our Wharton AI and the Future of Work Conference on 5/21-22. Last year was a great event with some of the top papers on AI and work. Paper submission deadline is 3/3. Come join us! Submit papers here: forms.gle/ozJ5xEaktXDE...forms.gle 21515
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 29/01/2025Could much of the recent drama about DeepSeek be rooted on flawed causal thinking i.e. neglecting the counterfactual? Dario Amodei makes a compelling case here. darioamodei.com/on-deepseek-... 231
Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 29/01/2025"Employment dynamics in a rapid decarbonization of the US power sector" Cool complexity economics analysis of a scenario that sadly looks a bit sci-fi in the current political circumstances. It would be interesting to study labour market impacts of AI with this approach. 121
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 29/01/2025"Employment dynamics in a rapid decarbonization of the US power sector" Cool complexity economics analysis of a scenario that sadly looks a bit sci-fi in the current political circumstances. It would be interesting to study labour market impacts of AI with this approach. 121
Reposted by Juan Mateos-GarciaTim Bresnahan @timobres.bsky.social · 27/01/2025Hello #EconSky, a simple question but one where I have not yet found the answer. What should I read to get a quantitative empirical understanding of the contemporary relationship, both size and timing, between scientific discovery and productivity growth? 41612
Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 27/01/2025"Suspense and surprise in technological innovation" Neat analysis of the link between expectations and outcomes in technological evolution. 153
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 27/01/2025"Suspense and surprise in technological innovation" Neat analysis of the link between expectations and outcomes in technological evolution. 153
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 25/01/2025Charles Dickens on Victorian UBI (from Bleak House 120
Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 24/01/2025"How technoscientific knowledge advances" The genesis of TS revolutions was one of the best innovation books in recent years. In this new essay, the authors summarise and expand its ideas. 152
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 24/01/2025"How technoscientific knowledge advances" The genesis of TS revolutions was one of the best innovation books in recent years. In this new essay, the authors summarise and expand its ideas. 152
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 17/01/2025Harnessing AI to Tackle Antimicrobial Resistance. Systematic assessment of the AI x AMR innovation system by the Fleming Initiative and Google DeepMind colleagues. We need more assessments like this in other areas of AI x science / social impact! www.imperial.ac.uk/Stories/harn...imperial.ac.ukHarnessing Artificial Intelligence to Tackle Antimicrobial ResistanceAntimicrobial resistance (AMR) is one of the most significant threats to global health security. The decade ahead will be critical in developing the necessary resilience to prevent and mitigate the b... 062
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 14/01/2025I am enjoying the Justified Posterior podcast, where Seth Benzell and Andrey Fradkin discuss economics of AI papers. They weren't that impressed with Acemoglu's macroeconomics of AI paper. podcasts.apple.com/gb/podcast/j...podcasts.apple.comJustified PosteriorsBusiness Podcast · 3 Episodes · Updated fortnightly 060
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 20/12/2024Christmas reading/re-reading ambitions. 060
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 19/12/2024AI is going to transform how I engage with theoretical economics papers. 170
Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 17/12/2024Clio: Privacy-preserving insights into Real-World AI use. Anthropic have developed a pipeline to cluster and visualise Claude user conversations. This is informing trust and safety efforts and creating evidence about AI real-world adoption. Very cool! assets.anthropic.com/m/7e1ab885d1... 1103
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 17/12/2024Clio: Privacy-preserving insights into Real-World AI use. Anthropic have developed a pipeline to cluster and visualise Claude user conversations. This is informing trust and safety efforts and creating evidence about AI real-world adoption. Very cool! assets.anthropic.com/m/7e1ab885d1... 1103
Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 10/12/2024The GenCast paper is very cool. I specially loved the analysis of the relative economic value of different weather prediction models. This shows that GenCast could improve efficiency by helping decision-makers prepare for extreme weather events when they need to. www.nature.com/articles/s41... 043
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 10/12/2024The GenCast paper is very cool. I specially loved the analysis of the relative economic value of different weather prediction models. This shows that GenCast could improve efficiency by helping decision-makers prepare for extreme weather events when they need to. www.nature.com/articles/s41... 043
Reposted by Juan Mateos-GarciaMel Andrews @bayesianboy.bsky.social · 29/11/2024So excited to see policy-oriented scholarship from researchers at DeepMind incorporating my work, and insights from philosophy of science more broadly. 1241
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 29/11/2024AI could deliver its biggest impacts by accelerating science and making new ideas easier to find. Our new essay draws on interviews and existing evidence to explore its opportunities, production function, risks and implications for policy. Summary thread: 030
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 27/11/2024Opacity and intelligence via Cormac McCarthy. 010
Reposted by Juan Mateos-GarciaTom Broekel @tbroekel.bsky.social · 27/11/2024Highly recommended and relevant. #GeoInnoSky 021
Reposted by Juan Mateos-GarciaBen Johnson @ersatzben.com · 27/11/2024I really liked the Google DeepMind paper on science for AI. There is a lot in there that was new to me. It is long, yes, but the encyclopaedic treatment -grounded in real examples- brings detail and colour to a discussion that suffers from under- and overhype. 182
Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 26/11/2024AI for science could be more impactful than chatbots. It is already helping win Nobel prizes and accelerating drug development and materials discovery. Today we published an essay about it: why it matters, how it’s happening and its implications. Here is a summary from an econ / social sci lens. 27930
Reposted by Juan Mateos-GarciaVivek Kumar @v1vekkumar.bsky.social · 26/11/2024Great 🦋-storm highlighting the importance and implications of AI for science. It's already helping win Nobel Prizes! 🤯 Juan's amazing essay to spark debate about the most important AI for Science opportunities, ingredients, risks and policy ideas. www.aipolicyperspectives.com/p/a-new-gold...aipolicyperspectives.comA new golden age of discoverySeizing the AI for Science opportunity 021
Reposted by Juan Mateos-GarciaSimone Vannuccini @svannuccini.bsky.social · 26/11/2024Great thread (and post) by @jmateosgarcia.bsky.social summarising how #AI can impact #science (when the AI bubble will burst or turn darkish, that's IMHO the real application that will survive)! As a true SPRUers, Juan gives a socio-technical&econ twist to the theme that is really needed! 021
Reposted by Juan Mateos-GarciaJohn Whitfield @johnwhitfield.bsky.social · 26/11/2024There's a lot on AI-driven discovery about at the moment, unsurprisingly (ht @maxnathan.bsky.social) 011
Reposted by Juan Mateos-GarciaMax Nathan @maxnathan.bsky.social · 26/11/2024Lots of brainfood in this excellent thread from Juan 👇 121
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 26/11/2024AI for science could be more impactful than chatbots. It is already helping win Nobel prizes and accelerating drug development and materials discovery. Today we published an essay about it: why it matters, how it’s happening and its implications. Here is a summary from an econ / social sci lens. 27930
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 21/11/2024“Should science by some miracle forge on into the future it will uncover not only new laws of nature but new natures to have laws about” (Cormac McCarthy via Bobby Western, The Passenger) 010
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 21/11/2024Nice paper about the epistemology of AI-driven Science illustrated with the case of AlphaFold. philsci-archive.pitt.edu/24182/ 24411
Reposted by Juan Mateos-GarciaBrett Karlan @oldjerryfodor.bsky.social · 20/11/2024GUY WHO DOESN'T GET AS MANY LIKES HERE AS HE DID ON TWITTER: I'm happy to share my new preprint paper, "Bluesky as Epistemic Injustice" 212820
Reposted by Juan Mateos-GarciaJuan Mateos-Garcia @jmateosgarcia.bsky.social · 24/10/2023"And you can love the internet and Artificial intelligence without romanticizing it, because it has phenomenal possibilities. It's extraordinary. But at the same time, be vigilant." I agree with Werner Herzog's AI policy position (from this interview: app.podscribe.ai/episode/8926...) 032
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 24/10/2023"And you can love the internet and Artificial intelligence without romanticizing it, because it has phenomenal possibilities. It's extraordinary. But at the same time, be vigilant." I agree with Werner Herzog's AI policy position (from this interview: app.podscribe.ai/episode/8926...) 032
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 20/10/2023Sociotechnical safety evaluations of generative AI systems. It was great to contribute to this GoogleDeepMind with Laura Weidinger and many other brilliant co-authors. Here are some reflections from an economics lens (#econsky). www.deepmind.com/blog/evaluat...deepmind.comEvaluating social and ethical risks from generative AIGenerative AI systems are already being used to write books, create graphic designs, assist medical practitioners, and are becoming increasingly capable. To ensure these systems are developed and depl... 111
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 19/10/2023Sociotechnical Safety Evaluation of Generative AI Systems New DeepMind paper proposing a holistic framework to evaluate generative AI models. I really enjoyed contributing to the section on post-deployment (layer 3) evals & evals of economic impacts. arxiv.org/pdf/2310.119... 000
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 18/10/2023"It is much as we expected, though the import of even the most expected news changes when it becomes definite, and fact." (Ship "Empiricist" via Iain M. Banks, to be thrown back at anyone complaining about unsurprising findings) 000
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 18/10/2023Can we predict the direction of AI research? Cool benchmarking of methods to predict how AI knowledge grows. Interestingly, methods relying on handcrafted network features beat end-to-end AI methods. www.nature.com/articles/s42...nature.comForecasting the future of artificial intelligence with machine learning-based link prediction in an ...The number of publications in artificial intelligence (AI) has been increasing exponentially and staying on top of progress in the field is a challenging task. Krenn and colleagues model the evolution... 100
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 17/10/2023John von Neumann defined GPTs (as well as building them) 000
Reposted by Juan Mateos-GarciaCarl T. Bergstrom @carlbergstrom.com · 16/10/2023Andreessen somehow always manages to stay in the running for worst dude in Silicon Valley. 158818
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 16/10/2023Could AI be used to produce scientific policy advice? According to this article, AI could summarise the literature and personalise outputs. Risks from hallucinations, biases and misinformation need to be mitigated with careful governance & humans in the loop. www.nature.com/articles/d41... 000
Juan Mateos-Garcia @jmateosgarcia.bsky.social · 12/10/2023Ethics of mathematics (via Labatut’s “The Maniac”) 000