Causality for Impact -> Dec 2025

Mark van der Laan learned one thing early in his career... ...and it helped him solve real-world problems: The structure of the problem matters. Mark is known as the godfather of Targeted Maximum Likelihood Estimation (TMLE). 1/ #CausalSky @EpiSky #StatSky #MLSky

What are the best books on causality and causal ML you've ever read? Why these ones? In 2018, I started reading everything on causality I could possibly find. At that time I had a few successful large-scale ML projects on my account both in business and academia. 1/ #CausalSky #MLSky #StatSky

This is one of the best real-world examples for why causality matters that I encountered so far. And it's extremely simple. 1/ #CausalSky #DecisionMaking #Business

💯

I just ordered a copy of this new book (vist.ly/46sv8) When we think about complex systems, we often think about non-linearity, self-organization and emergence. In the context of causality, emergence is often of special interest. 1/ #CausalSky

He did it before Double Machine Learning I met with professor Mark van der Laan because I think his work is pretty incredible and it sometimes feels like a secret that only a few people know about, especially in industry. 1/ #CausalSky #StatSky #CausalInference

Hey folks, I'm looking for references on the use of information geometry in causal inference/discovery. Have you come across any papers like this? #CausalSky #StatSky #MachineLearning

33% of people reported having 0 people among their colleagues who understand causality. What's the impact of this fact on the business they work for? That's a causal question, and hence their colleagues likely cannot answer it. 1/ #CausalSky #CausalInference

Does probability exist? Over the last two weeks, David Rohde shared his thoughts on realism and anti-realism in statistics and causal inference in our newsletter (part 2: vist.ly/45ywf). 1/ 👇🏼 #CausalSky #StatSky #Probability

Propensity Score Matching (PSM) can mislead you badly. Here are 3 caveats on PSM, and the third one is the trickiest: 1️⃣ I learned that PSM is often taught as if it was a magic method that can recover causal effects from arbitrary observational data. 1/ 👇🏼 #CausalSky #StatSky

A Brand-New Book on Causal AI (FREE) Elias Bareinboim is an Associate Professor of Computer Science at Columbia University and the director of the CausalAI Lab. 1/n #CausalSky #CausalAI

I'll send your friend a book. When I visited my parents over the weekend, my father showed me this: It’s his collection of every available translation of my book. He orders each one from... 1/ 👇🏼 #FreeBook #Causality #CausalAI #CausalSky

Can you peek at p-values in your A/B test? You can, but you need to be SAVI. Common sense teaches us that peeking at p-values in our A/B tests (or other types of trials) and making decisions to stop the experiment based on what we see is a grave mistake. 1/ 👇🏼 #CausalSky #ABTesting #StatSky

Have you ever thought about the problem of multiple testing in the context of sequential decision-making? Whether you answered yes or no, this talk is for you (especially if sequential problems spark your curiosity). 1/ 👇🏼 #CausalSky

Models and Decision Rules Are Not the Same (And Why It Matters for Causality) (We'll start sending it out at 12PM ET / 6PM CET today) In this week's issue: 1/ 👇🏼 #CausalSky

Nice!

Happy to present on causal dynamic predictions at #ISCB46 in Basel! Join me, @gelovennan.bsky.social‬, @vanamsterdam.bsky.social, and @ruthkeogh.bsky.social‬ for the invited session “Prediction modelling meets causal inference for clinical decision making”. 📍 Mon 25 Aug, 16:00 (Biozentrum U1.111)

Hey folks, Aug 26 -> Sequential causal inference in experimental or observational settings #CausalSky Registration: www.niss.org/events/seque...

Better Roll-out Decisions with Your A/B Tests? Online platforms often run many, sometimes thousands, of randomized online experiments (A/B tests) in parallel. One of the challenges, even for big platforms, is... 1/ 👇🏼 #CausalSky #ABTesting

Pleased to report on two papers on #ANVOVA with @poschm.bsky.social and Franz König The first lnkd.in/eKmjiQdk looks at median stratification for a single covariate. The second lnkd.in/e3DH96G8 considers adjustment for many covariates. Variance inflation factors are key.

The Horvitz–Thompson Estimator Has a Problem. David Rohde explains why, and what to use instead. (This week in causal inference--two perspectives worth your attention) 1/ 👇🏼 #CausalSky

You (probably) should not think about price elasticity this way. This week, I met with Łukasz Frydrych & Artur Machno to work on a pricing problem to help one of the Fortune 500 companies. 1/ 👇🏼 #CausalSky

🎉 Last week Kamyar Barakati received the 1st prize in the poster competition at Microscopy & Microanalysis, presenting our work on LLM-assisted causal discovery. Congrats Kamyar & the entire team! 1/2 #CausalSky

Tomorrow, my colleagues will present our work on LLM-assisted causal discovery for mapping hidden patterns in crystalline solids. If you're in Utah, come and visit them! #CausalSky

"We had 90% of people who 'agreed' or 'totally agreed' to the statement that the session was 'fruitful and constructive'. That's among the highest numbers we ever had on this event or other trainings." 1 /👇🏼 #CausalSky

"Recommender system is NOT a causal problem" Or is it? A short answer is that "it depends". A longer one is that if you care about optimizing your business metrics, then you will probably be much better off if you consider recommendation a causal problem. 1/ 👇🏼 #CausalSky

In the last issue of Causal Python Weekly, Alberto D. Horner reviewed Re-Imagine: A Causality-Inspired Framework to Measure Generalizable Reasoning. As Alberto writes: 1/ 👇🏼 #CausalSky #AI

Fully agree @p-hunermund.com :) #CausalSky

Benchmark Me a River: A New Causal Discovery Benchmarking Kit is Here Causal discovery from observational data is often deemed one of the most challenging tasks in causal modeling. The limited availability of real-world causal discovery benchmarks doesn't make it any easier. 1/ 👇🏼 #CausalSky

New paper alert: Opportunities for Causal Machine Learning in Precision Oncology (NEJM AI) Can Causal ML assist us in selecting best personalized treatments? 1/ 👇🏼 #CausalSky #PersonalizedMedicine #MachineLearning #AI

Want more power for CATE in your A/B test? Imagine that you're running a randomized controlled trial (RCT/AB test) and, subsequently, you want to use the data from this experiment for personalized decision-making. 1/ 👇🏼 #CausalSky #StatSky

Hidden confounding is one of the top issues listed by practitioners in observational and mixed-data causal inference. While frameworks like causal sensitivity analysis and partial identification are slowly getting traction in the community, ... 1/ 👇🏼 #CausalSky

@rickardkarlsson.bsky.social has some exciting new papers at ICML! #CausalSky

Time to have some fun with planning! Can causality and physics help models surpass human performance? Tasks involving complex interactions between objects with unknown dynamics present significant challenges for planning before execution. 1/ 👇🏼 #CausalSky

It’s often so much easier to focus on obstacles rather than work on creative solutions. Hidden confounding? Uncertainty? Embrace them and… …give them a place where they belong. They are part of the process. 1/ 👇🏼 #CausalSky

In case you're confused, there's an arrow indicating the direction of the tail. #AI #StatsSky

Should we forget about model-free general agents? IMHO, a must-read for anyone who's serious about agentic systems. One of the differentiating factors of human intelligence, as compared to the current generation of AI systems, is the ability to perform... 1/ 👇🏼 #CausalSky

Is prediction and decision-making the same? On Friday, I'll be in London talking about effective decision-making and why causal thinking is a necessary ingredient in the process. #CausalSky 1/ 👇🏼

Andersen's "The Emperor's New Clothes" final scene ends with a child blurting out what would be obvious to any outsider: "The king is naked!" The tale, based on a medieval Spanish story from "Libro de los ejemplos" ("Book of Examples"), illustrates potential consequences of... 👇🏼 1/ #causalSky

Changing the perspective Low code time :)

Throwback: Difference-in-Differences: A Practitioner's Guide The Dream Team's Perspective Difference-in-Differences (DiD) is arguably the most popular quasi-experimental research design. 1/ 👇🏼 #CausalSky

A quick reminder: this course is free for everyone. Fun fact: more people enroll on Mon and Sun than any other day of the week. We discuss causality in the context of experimentation and modern machine learning. 400+ people took it, and most think it's a 5-star course. 1/ 👇🏼

Submissions for this year's edition of Causal Data Science Meeting are open! Submission deadline: September 30, 2025. This year we'll hear Stefan Feuerriegel (LMU) as the keynote speaker. As the organizers say: ... 1/ 👇🏼 #CausalSky

Hey folks, if anyone's interested, there are still a few hours to submit your papers for Royal Stat Society's event on innovative usages of natural experimewnts: rss.org.uk/news-publica... #CausalSky #StatSky

I once told my wife that I can talk about causality for hours. I didn't mean it figuratively, though, LOL. This week we ran a workshop on causality for my banking client that started at 8AM and finished at 6PM. 1/ 👇🏼 #CausalSky

Is Jonathan Heidt right that social media causes the mental health crisis in young people? If so, how can we be sure? Can other disciplines learn something from the reproducibility crisis in Psych & what is multiverse analysis? Join me & @dingdingpeng.the100.ci‬ for a conversation on... 1/👇🏼

Not all real-life decisions are about one dimensional outcomes. When applying for a new job, we might care about the compensation, but also the company's mission or their work-from-home policy. 1/ 👇🏼 #CausalSky

Some interesting work suggesting that recent shocking RL+LLM results are due to incorrect baselines and most of the gains are from better format following: safe-lip-9a8.notion.site/Incorrect-Ba...

@alisongopnik.bsky.social on intelligence