iiinova — where your digital self keeps evolving

Charlie Rahal @crahal.com · 24/09/2026

We started a data science consultancy company to run alongside our academic work. We have a limited amount of time that we can devote to this, but please do get in contact if you think there's anything we can help with! patterns-and-proofs.com

112

Charlie Rahal @crahal.com · 04/09/2026

Closing slides from this year’s fabulous 7th International Conference on Social Computing 🎉 Huge thanks to @nuffieldcollege.bsky.social for hosting: this was incredible fun, with fascinating frontier research presented. tinyurl.com/3nask76e See you all next year in Chengdu 🍜🐼📈 !

100

Reposted by Charlie Rahal

Neil Renic @ncrenic.bsky.social · 31/05/2025

GPT-6 will have the intelligence to open a bakery instead of doing a PhD

5486197

Charlie Rahal @crahal.com · 28/08/2026

New work \w @sanderwagner.bsky.social and @melindacmills.bsky.social🔬🔭. LLM-based extraction × classification × unstructured REF narratives × large-scale scientometrics 📚. Headline: Women more represented in research underpinning impact than in research output itself 👩‍🔬: arxiv.org/abs/2608.26409

030

Charlie Rahal @crahal.com · 20/08/2026

How reliable is research 🤔? We publish RobustiPy in @cp-patterns.bsky.social: a highly accessible, #GNU-GPL 3.0 #python lib for #robustness analysis and #datascience. It may represent the “next generation” of #modeluncertainty, bringing a range of frontier #statistics methods into one toolkit 📊📈!

1113

Charlie Rahal @crahal.com · 10/08/2026

The GWAS Diversity Monitor is now firing on all cylinders—and running faster than ever🚀⚡! All the relevant GitHub issues have been officially crushed💥✅. A huge thank you to Mingyue and Daniel for the fantastic teamwork that made it happen👏🫶🎉! Take a look: gwasdiversitymonitor.com 🔎🧬

030

Charlie Rahal @crahal.com · 06/08/2026

Today at Metrics and Models, Jacob Walker talks about "Computer Vision Research at Google DeepMind"💻👀. Should be good, all welcome! Register via deets on our homepage ⬇️: metrics-and-models.github.io or send a DM for a link. Next up is @pengzell.bsky.social talking about "The Paper Factory" 🏭📝!

020

Charlie Rahal @crahal.com · 04/08/2026

We'll shortly be advertising *four* of our doctoral projects for 2027 entry 🎉📢 (stay tuned 👀)! If you have other projects in applied health and social data science that you'd like me to point people towards, let me know! Happy to help amplify, or even form a small colloquium of resharing 🤝❤️💬.

033

Reposted by Charlie Rahal

Pat Hastings @ophastings.bsky.social · 21/07/2026

Feeling deprived of academic talks over the summer? This Thursday I'm presenting new work with @zparolin.bsky.social for the Oxford Metrics & Models seminar. Online talk and all are welcome! 2pm Oxford time, 7am Colorado time. Happy to share the MS Teams link with you. metrics-and-models.github.io

Talk Title: Rags or Riches? Predicting Life Outcomes from the Birth Lottery Across Five High-Income Countries

Abstract: Stratification in life outcomes is a central concern of sociology, but to what extent can adult outcomes be predicted from childhood circumstances? Although decades of research have identified important family and socioeconomic factors associated with mobility and attainment, these models explain only a modest share of individual variation in outcomes. Using harmonized longitudinal data from the Cross-National Equivalent Files (CNEF) and an ensemble of machine-learning models applied to a common set of childhood characteristics, we estimate the predictability of adult outcomes in the United States, United Kingdom, Germany, Australia, and South Korea. Predictability varies substantially across outcomes and national contexts: educational attainment is the most predictable in general, poverty is far more predictable in the U.S. than other countries, and income mobility is substantially less predictable than overall attainment outcomes. In additional analyses with the U.S., adding more variables including race, wealth, and parental incarceration improves predictions only slightly. We further show that predictability captures a distinct dimension of stratification that is not fully captured in other traditional measures of mobility, inequality, and opportunity. Together, these findings provide new insights into where and how the birth lottery most strongly shapes life chances.

Bio: Pat Hastings is an Associate Professor of Sociology at Colorado State University. He studies parenting, childhood, inequality, and social mobility with quantitative and computational methods.

1124

Charlie Rahal @crahal.com · 19/05/2026

CW: suicide. Honoured to be Guest Editing—with @saurabhkhanna.bsky.social—a Special Issue at ACM-TSC in tribute of my intellectual hero, Aaron Swartz, who died by suicide in 2013 at 26 after extreme prosecutorial overreach. While attempting to expand access to knowledge. CfP: tinyurl.com/4znbk4y2.

122

Charlie Rahal @crahal.com · 11/05/2026

Two weeks left until our Call for Papers closes for the 7th International Conference on Social Computing (ICSC'26). Please do submit your papers! Optional conference proceedings, great keynote speakers, nice conference dinner @nuffieldcollege.bsky.social, and more. Details: icsc-conf.github.io!

094

Charlie Rahal @crahal.com · 04/05/2026

Today's pushes bring in some speed efficiencies to our new model uncertainty library which make it approximately **14** times faster! Check it out now at robustipy.github.io; accompanying paper getting published very soon, super jazzed 🎷

062

Charlie Rahal @crahal.com · 05/04/2026

Delighted to be helping chair ICSC '26 🤖. Hosted: @nuffieldcollege.bsky.social 🏰 When: Sept ’26 🕰️ Details: CfP live 🎉 Optional proceedings, all welcome. More: icsc-conf.github.io. Please consider sharing your work📜! Kindly help us spread the word, and follow @i-c-s-c.bsky.social for updates📣!

1104

Charlie Rahal @crahal.com · 24/12/2025

The quote of the year came late, but it’s good: “This is LLM generated nonsense. You make a mockery of real research. Do not submit this again.” If the originator of the quote sees this, you will remain entirely anonymous I assure you, but I want you to know that I appreciate you. Happy holidays!

0517

Reposted by Charlie Rahal

Leverhulme Centre for Demographic Science, Oxford University @oxforddemsci.bsky.social · 24/11/2025

📖👨‍💻🎓 New Courses Alert 🎓👨‍💻📖 The Leverhulme Centre for Demographic Science is launching three upcoming short courses designed for researchers across health, economics, and the social sciences. Led by experts in the field — including @crahal.com — these courses are now open for registration.👇

165

Reposted by Charlie Rahal

Leverhulme Centre for Demographic Science, Oxford University @oxforddemsci.bsky.social · 14/11/2025

A pleasure to host @philipncohen.com this week for a fantastic talk on pronatalism & many more engaging discussions about open science, public engagment, and more! #demography @nuffieldcollege.bsky.social @drjenndowd.bsky.social @melindacmills.bsky.social @crahal.com @ridhikashyap.bsky.social

1223

Charlie Rahal @crahal.com · 11/10/2025

We're holding brief/extremely informal 15 minute meetings with anyone who is interested in either of these positions, scheduled for 19.11.2025 and 17.11.2025, respectively. Email or DM for more information! Please do share details within your network; I think these projects are pretty cool!

078

Charlie Rahal @crahal.com · 01/10/2025

Thrilled to serve as TPC co-chair (/w @chenyang03.bsky.social) for the brilliant International Conference on Social Computing! 🚀 See @i-c-s-c.bsky.social for updates! 📅 Papers: October 20th! 🔗 Full details: icsc-conf.github.io/2025/ Let’s push the frontiers of #CSS, #DataScience, #LLM, and #AI💡✨

030

Charlie Rahal @crahal.com · 29/09/2025

🎉Our metrics-and-models.github.io lab announces first hybrid seminar in conjunction with @oxforddemsci.bsky.social🚨 "Nonparametric Causal Decomposition of Group Differences: New Mechanisms & New Methods" Speaker: Felix Elwert When: October 8th, 2025 Where: @nuffieldcollege.bsky.social

061

Charlie Rahal @crahal.com · 18/09/2025

A paper of ours just got rejected from a leading computer science journal for being ‘too technical’ and for some disturbing reason the inner economist in me is really proud.

060

Charlie Rahal @crahal.com · 05/09/2025

The utility of this data resource cannot be understated. Infinite respect to all of the incredibly talented people who worked on it! 😍

020

Charlie Rahal @crahal.com · 01/09/2025

Thanks for asking! Its almost like we made a way to automatically track this on a daily basis, @pietrobiroli.bsky.social 🤩! Kindly see gwasdiversitymonitor.com. See also: Mills, M. C., & Rahal, C. (2020). The GWAS Diversity Monitor tracks diversity by disease in real time. Nature genetics, 52(3).

gwasdiversitymonitor.com

GWAS Diversity Monitor

This interactive dashboard monitors participants included across all published Genome Wide Association Studies (GWAS).

240

Charlie Rahal @crahal.com · 01/09/2025

🚨It's that exciting time of year!🚨 Our team has got TWO open DPhil (PhD) positions in #datascience + #health here at @oxforddemsci.bsky.social for 2026🎉! See for details: 1️⃣ 👉 crahal.com/HDS_2026.html 2️⃣ 👉 crahal.com/pophealth_2026.html Please share and help us to find our future teammates ❤️!

12617

Charlie Rahal @crahal.com · 27/08/2025

💡🤖🔥 @keyonv.bsky.social's talk at metrics-and-models.github.io was brilliant, posing epistemic questions about what Artificial Intelligence "understands". Next (two weeks): Alexander Vezhnevets talks about a new multi-actor generative agent based model. As usual, *all welcome* #datascience #css💡🤖🔥

041

Charlie Rahal @crahal.com · 27/08/2025

Excited to contribute to this; very exciting stuff by @hggaddy.bsky.social, as usual!

170

Charlie Rahal @crahal.com · 14/08/2025

💡🤖🔥The talk by Juan Carlos Perdomo at metrics-and-models.github.io was so thought provoking that the convenors stayed to discuss it in the room afterwards for quite some time! Next, we have @keyonv.bsky.social asking: "What are AI's World Models?". Exciting times over here, all welcome!💡🤖🔥

062

Charlie Rahal @crahal.com · 04/08/2025

💡🤖🔥The talk by @klint.bsky.social at metrics-and-models.github.io was truly brilliant. Important work; a clean/intuitive computational design & empirical application from the @itemrespwarehouse.bsky.social! *Next:* Juan Carlos Perdomo on "The relative value of prediction" (v. excited about this)💡🤖🔥

130

Charlie Rahal @crahal.com · 17/07/2025

@ramina-sotoudeh.bsky.social gave an AMAZING talk yesterday as part of the #metrics-and-models seminar series👏. Next (30.07), we have @klint.bsky.social on "Reproducing Expert Judgement with Shortened Surveys using Simulated Annealing"🔥! See website for more; metrics-and-models.github.io!

021

Charlie Rahal @crahal.com · 02/07/2025

The first #metrics-and-models talk was fantastic🤗! Next: @ramina-sotoudeh.bsky.social on "Antenatal Selection: A novel approach for assessing the genetic and ecological drivers of who survives to birth". Details: metrics-and-models.github.io Sign up: metrics_and_models-subscribe@maillist.ox.ac.uk

0104

Charlie Rahal @crahal.com · 30/06/2025

Hurrying to post about RobustiPy, i forgot this fun animation i made for teaching purposes (with @deevybee.bsky.social on the SABS R3 CDT) around when we started developing. Specification uncertainty is just _one_ type of uncertainty, though; our forthcoming work looks at the role of random numbers!

061

Charlie Rahal @crahal.com · 30/06/2025

🚀 RobustiPy is out of beta! Feature rich, we hope it becomes useful for model uncertainty / #datascience. We include 🔟 replications, including “unexplained discrepancies” in prior work! 📄 Preprint: arxiv.org/abs/2506.19958 💻 GitHub: github.com/RobustiPy/robustipy 🌐 Homepage: robustipy.github.io

arxiv.org

Introducing RobustiPy: An efficient next generation multiversal library with model selection, averaging, resampling, and explainable artificial intelligence

We present RobustiPy, a next generation Python-based framework for model uncertainty quantification and multiverse analysis, released under the GNU GPL v3.0. Through the integration of efficient boots...

1137

Charlie Rahal @crahal.com · 10/06/2025

🔥Coming July 2: 'Metrics and Models'!🔥 A new seminar series on advanced modelling w/ impact across health/society. 🌍 Open to *ALL*! 🕑 Weds 2pm UK alternating weeks (online) 🔗 metrics-and-models.github.io 📬 metrics_and_models-subscribe@maillist.ox.ac.uk Brilliant speakers. *Please share*!

12617

Charlie Rahal @crahal.com · 15/05/2025

“it’s quite heavy on the statistical analysis” is probably one of the nicest inadvertent compliments a politician could ever give to a data scientist. Thank you kindly, @wendychambld.bsky.social!

030

Charlie Rahal @crahal.com · 11/04/2025

One of the world's great books now accompanied by one of the world's great programming languages ♥️. Can't wait to read this.

010

Charlie Rahal @crahal.com · 08/04/2025

Just moved to @zed.dev as my go-to text editor. Blazing fast with minimum latency (of any sort), so extremely feature rich, immensely beautiful, and clearly an IDE made by people who know what developers want and need. I absolutely love it; highly recommended!

2292

Reposted by Charlie Rahal

Christopher Barrie @cbarrie.bsky.social · 30/03/2025

Best guides people know for updating pre registration on the basis of pilot or first wave experiments?

063

Charlie Rahal @crahal.com · 27/03/2025

1/5 📄✨ITS FINALLY HERE!📄✨. After _five_ years, our new model evaluation metric — The InterModel Vigorish (IMV) — is now published (@plosone.org). Paper: bit.ly/3RlqrQH Preprint: bit.ly/41LzGjm Code: github.com/intermodelvigorish ‘What’s the vig on this action?’ (bit.ly/4hPoZ3Z). ⬇️⬇️⬇️

1126

Charlie Rahal @crahal.com · 15/03/2025

🎉🚀🔬Our new model uncertainty library -- RobustiPy -- now also has joint inference tests (as per @urisohn.bsky.social et al., 2019, NHB). Near to release/preprint! GitHub: github.com/RobustiPy/robustipy project website: robustipy.github.io/ /w @dvaldenegro.bsky.social & @jiani-y.bsky.social!🎉🚀🔬

0102

Charlie Rahal @crahal.com · 08/03/2025

📖🎲RROx is hosting a fabulous Open Scholarship version of the 'Would I Lie to You' gameshow as part of the Oxford Forum of Open Scholarship.📖🎲 Some great #openscience stories to tell! Weston LIbrary (Oxford) at 15:30 on 12.03.2025, still some spaces left 🙃. Signup: shorturl.at/5EB2S

020

Charlie Rahal @crahal.com · 04/03/2025

📢🎉🤖 Published today in natcomputsci.bsky.social /w @jiani-y.bsky.social📢🎉🤖: A new framework for (ir)reducible errors. We argue that it's impossible to know how 'predictable' things are (until we've collected every epsilon of information). ReadCube: rdcu.be/eb9IL arXiv: arxiv.org/abs/2411.19223

074

Charlie Rahal @crahal.com · 24/02/2025

In an age when so much modern scientific conversation surrounds 'Artificial General Intelligence', lets not forget that René Descartes -- in De homine (1662) -- claimed that non-human animals could be explained reductively as automata.

061

Charlie Rahal @crahal.com · 24/02/2025

📝Registration for events at the Oxford Forum of Open Scholarship still open to all (March 3rd-13th, 2025)!📝 Great speakers like @briannosek.bsky.social & many others👀; looking forward to chairing a @pkp.sfu.ca / @elife.bsky.social session, too! Registration/Schedule: ⬇️ openaccess.ox.ac.uk/oxfos25

082

Reposted by Charlie Rahal

Malika Ihle @malikaihle.bsky.social · 15/01/2025

Meta-science postdoc position (3yrs): Agent-based modeling of academia 📍 Munich, Germany (Remote/hybrid options may be available) 🗣️ Felix Schönbrodt @nicebread.bsky.social, director of @lmu-osc.bsky.social 📅 application: 21.01.2025, start: March-July 2025 www.psy.lmu.de/pm/aktuelles...

psy.lmu.de

Meta-science postdoc position (3yrs): Agent-based modeling of academia - Lehrstuhl für Psychologische Methodenlehre und Diagnostik - LMU München

05435

Charlie Rahal @crahal.com · 25/12/2024

“you will be visited by three types of epistemic error” arxiv.org/abs/2411.19223

061

Reposted by Charlie Rahal

Professor Melinda Mills @melindacmills.bsky.social · 18/12/2024

Great opportunity to work with @crahal.com on his frontier informatics ideas & join @oxforddemsci.bsky.social

083

Charlie Rahal @crahal.com · 17/12/2024

📝🚨Disruptive Preprint Alert🚨📝! @jiani-y.bsky.social and I conceptualise aleatoric/decompose epistemic error in generalised predictive systems, claiming that you can't ever really know the limits of #ai #ml accuracy. Keen for thoughts; kindly share, we'll acknowledge! URL: arxiv.org/abs/2411.19223

arxiv.org

On the Unknowable Limits to Prediction

This short Correspondence critiques the classic dichotomization of prediction error into reducible and irreducible components, noting that certain types of error can be eliminated at differential spee...

031

Charlie Rahal @crahal.com · 10/12/2024

Now on arXiv! We design a new two-stage Individual Synthetic Control algorithm to cleanly identify causal estimates of the “caring penalty.” High-intensity unpaid carers face ~45% personal losses; households lose ~12%. 🏠📉 Link: arxiv.org/abs/2411.10314 Blog: shorturl.at/IVQ6M

02710

Charlie Rahal @crahal.com · 02/12/2024

🌟🚨🔥We’re hiring a postdoc!🔥🚨🌟 Got skills or interest in: 🔹 NLP\LLMs? 🔹 Advanced Python dev? 🔹 How HPCs work? Join us @oxforddemsci.bsky.social and the @centreforcare.bsky.social to innovate! 📄 Apply: shorturl.at/c1obm ⏳ Deadline: Jan 13, 2025 Feel free to DM me w\ Qs, and kindly share widely!🙌🙌

045

Charlie Rahal @crahal.com · 22/11/2024

🚨Two Exciting DPhil\PhD Opportunities to join us at @oxforddemsci.bsky.social in Fall 2025!🚨 1️⃣ ABM project with 'Big Data' & LLMs (HDS CDT):👉 shorturl.at/wcMQo 2️⃣ Scientometrics project (DPhil PopHealth):👉 shorturl.at/qwsor 🕒⏳Deadline: ~10 days! Feel free to DM me w\ any Qs. Please share widely! 🙌

01415