Study Group – Week 1: From Prediction to Intelligent Action
On Thursday October 15, 2026, from 12:00–13:30 PM (UTC-6), or 18:00–19:30 PM (UTC), our Discord community will hold a study group meeting! This is a study group organised by members of our community. The aim of this study group is to share questions, solve exercises, and discuss thoughts on AGI. The focus of this next…
Dr. Bruno Bauwens on Solomonoff Induction of Selected Bits
In the next UAI research meeting on Oct. 12th at 3 pm EDT, Dr. Bruno Bauwens will present his work on Solomonoff induction of selected bits. We will use the regular meets link (also in the UAI Reading Group calendar‘s description). Bio Dr. Bruno Bauwens is an Associate Professor at the Higher School of Economics…
Mikhail Mironov on Stopping Complexity
In the next UAI research meeting on Sep. 28th at 3 pm EDT, AIXI labs fellow Mikhail Mironov will discuss his recent work on stopping complexity. We will use the regular meets link (also in the UAI Reading Group calendar’s description). Bio: Mikhail Mironov is an AI safety researcher and research manager with a PhD…
Failure of the coding theorem for randomized stopping machines
In a new blog-post on the lesswrong forum, AIXI labs fellow Mikhail Mironov explores the failure of the coding theorem for randomized stopping machines. This post explains a technical separation result in algorithmic information theory, which was derived during Mikhail Mironov’s Summer 2026 PIBBSS fellowship. The result contributes to AIXI Labs’ research program on how…
Michele Vannucci on Imitation Learning
Michele Vannucci, research fellow at AIXI labs, will discuss the paper “Imitation Learning is Probably Existentially Safe” by Cohen and Hutter at the weekly UAI research meeting today, Monday Sep. 21st at 3 pm EDT. We will use the regular meets link (also in the calendar description). Bio: Michele Vannucci recently graduated from a Master’s…
September 2026 Fellowships Awarded
The next two AIXI Labs fellows have been selected: Michele Vannucci and Ravi Deedwania. Congratulations to both! Other applicants will automatically be reconsidered for future fellowship rounds (the next starts in December).
Rubi Hudson on the Corrigibility Transformation
Rubi Hudson will talk about his AI safety research on the corrigibility transformation at the weekly UAI research meeting tomorrow, Monday Sept. 7th at 3 pm EDT. We will use the regular meets link (also in the calendar description). Bio: Rubi is a 5th year PhD student at the University of Toronto and research lead…
Iliad: If you understand AIXI, AI safety has work for you
This is a guest post by Iliad. Our thanks to Marcus Hutter for the invitation to post here. TL;DR: Iliad runs funded training and research programs in mathematical AI alignment, and we think the UAI community is well prepared for them. Applications for the October and November Intensives are open (apply here). If the programs…
Aram Ebtekar justifies Solomonoff induction on finite data
Aram Ebtekar, a returning speaker and founding research scientist at AIXI Labs, will present his (new) justification for Solomonoff induction at tomorrow’s regular research meeting, Monday July 13th at 3 pm EDT. Aram discusses the obstacle that No Free Lunch poses for developing a theory of generalization, and a practical sense in which Solomonoff induction,…
Anthony DiGiovanni on Safe Pareto Improvements
Anthony DiGiovanni, a researcher at the Center on Long-Term Risk, will talk to us about his Safe Pareto Improvements research agenda during an extra meeting tomorrow (Monday, July 6th), 3 pm EDT. We will use the regular zoom link. The Safe Pareto Improvements (SPI) agenda is intended to reduce the risks of catastrophic cooperation failures…
June 2026 Fellowships Awarded
The first two AIXI Labs fellows have been selected: Yegon Kim and Gabriel Leuenberger. Congratulations to both! Yegon continues his work on model-free AIXI, while Gabriel studies embedded agency and Occam’s razor. Other applicants will automatically be reconsidered for future fellowship rounds (the next starts in September).
Algorithmic Compression via Pretrained Neural Networks
Marcus Hutter describes this new paper as “a short recap of the last ~15 publications over 5 years of the Universal Artificial Intelligence team” (at Google DeepMind): https://www.mdpi.com/1099-4300/28/6/596 Abstract: The success of large neural networks trained for sequential prediction via log-loss minimization over massive and diverse datasets has sparked debate regarding the fundamental limits of…
Reading group: Kolmogorov Complexity and its Applications
The Australasian AIT reading group will be holding weekly meetings to discuss assigned readings from Li & Vitanyi’s An Introduction to Kolmogorov Complexity and its Applications, every Tuesday at 12:00 AEST (Monday at 19:00 PT). This is the canonical reference on the mathematical foundations of algorithmic information theory! No reading is assigned for the first…
Apply for a fellowship with AIXI Labs!
I am excited to announce that we are now offering fellowships at our new ASI safety org, AIXI Labs! Apply here: https://docs.google.com/forms/d/e/1FAIpQLSdpKSZKcMUZCkMaBncmbDSSdVcHKodDmdALUTOOheGF_NMHuQ/viewform?usp=publish-editor Focus. If you have been closely engaged with this community (or AIT, AIXI, and/or ML), you should consider applying! There is a good chance we can find a research match, since we are…
Francesca Zaffora Blando on Weak Merging of Opinions
Professor Francesca Zaffora Blando will present her work on weak merging of opinions at the regular research meeting tomorrow, Monday April 27th, at 3 pm EDT. Francesca Zaffora Blando is an Assistant Professor in the Department of Philosophy at Carnegie Mellon University specializing in logic, formal epistemology, and the philosophy of science – particularly the…
Sven Neth Against (Proxy) Optimization
Sven Neth will discuss his paper “Against Optimization” at an extra research meeting tomorrow, April 13th at 3 pm EDT, using the regular zoom link. This will be our second consecutive talk on AI safety (related topics)! Indeed, Sven’s paper was inspired by the limitations of “Consequences of Misaligned AI” by Zhuang and Hadfield-Menell, and…
Announcing the Third Symposium on AIT & ML!
We are excited to announce the Third Symposium on Algorithmic Information Theory and Machine Learning this July 27 – 29 at the University of Oxford! See the details here: https://sites.google.com/site/boumedienehamzi/third-symposium-on-machine-learning-and-algorithmic-information-theory The third iteration of the symposium is particularly focused on applications of AIT to the theory and practice of AI safety. AIXI has long been…
Aram Ebtekar on AIXI with Golden Handcuffs
Aram Ebtekar returns to give his second talk at the regular UAI research meetings this Monday (3 pm EDT on April 6th at the regular zoom link), this time on an exciting new AI safety protocol analyzed rigorously in the AIXI setting. Title: Golden Handcuffs make safer AI agents Abstract: Reinforcement learners often find novel…
Yegon Kim on Model-Free UAI
Yegon Kim will present tomorrow, Monday March 23rd at 3 pm EDT on a model-free universal artificial intelligence: https://arxiv.org/abs/2602.23242 We will use the regular zoom link. Bio: Yegon Kim is a PhD student researching AI at KAIST. Yegon also has an interesting blog/personal website here: https://yegonkim.github.io/ Abstract: In general reinforcement learning, all established optimal agents,…
Anders Christian Hansen on the Consistent Reasoning Paradox
Professor Anders Christian Hansen will discuss his work on the “Consistent Reasoning Paradox” at an extra research meeting tomorrow, Monday March 16th at 3 pm EDT. We will use the usual zoom link. I am a priori skeptical that logical paradoxes block AGI e.g. the Penrose-Lucas argument seems (especially) confused from the deep learning perspective,…
Tom Sterkenburg on Solomonoff Induction
Dr. Tom F. Sterkenburg will present a rigorous critique of Solomonoff induction at the UAI regular research meeting on Monday, March 9th, at 1:15 pm EST rather than the usual 3 pm EST. We will use the usual zoom link. Sterkenburg is Emmy Noether junior group leader at the Munich Center for Mathematical Philosophy (MCMP,…
Marta Kryven on Approximate Bayesian Inference for Problem-solving Domains
Marta Kryven will present on approximate Bayesian inference for problem-solving domains at the UAI research meeting tomorrow, Monday February 23rd at 3 pm EST. We will use the regular zoom link. Abstract: Bayesian Inference and Optimal Experiment Design provide a general case optimal theoretical framework for how to build world models and interpret evidence. However,…
Universal Search Reading Group Starts March 9th
Logistics We’re starting a reading group on universal search. See here for a list of readings we’re planning to go through. The reading group will meet weekly on Mondays starting on March 9th. The normal start time will be 1 PM EDT, but for the first session on March 9th we’ll start at 12 PM…
Elija Perrier on Quantum AIXI
Elija Perrier will discuss his work on quantum mechanical versions of the AIXI agent: https://arxiv.org/abs/2505.21170 The (extra) talk will be tomorrow, Monday Feb 16th at 3 pm EST over the usual zoom link. Bio Elija is a doctoral candidate at the Centre for Quantum Software and Information at the University of Technology Sydney. Elija’s doctoral…
Marc Finzi on Epiplexity
Dr. Marc Finzi will discuss his work on computationally bounded information theory (“epiplexity”) at this week’s regular research meeting, Monday February 9th at 3 pm ET (at the regular zoom link). Details below. Bio:Marc Finzi is a Research Scientist at OpenAI. He received his Ph.D. in Computer Science from New York University for work on…
Q&A with Marcus Hutter at the final meeting of the IUAI reading group
Marcus Hutter will answer any lingering questions on “An Introduction to Universal Artificial Intelligence” at the final meeting of the reading group on Monday, Jan 26th at 12 pm EST = 5 pm GMT. The zoom link is (as usual): https://uwaterloo.zoom.us/j/7921763961?pwd=TDatET6CBu47o4TxyNn9ccL2Ia8HN4.1 Note that the Australasian reading group is several chapters behind and still running.
Zhengmian Hu on Dialectics for AI
Zhengmian Hu is a researcher at Adobe Research. He received his Ph.D. from the University of Maryland, College Park, and has published across topics including watermarking, deep learning theory, and optimization theory, with an emphasis on theoretical foundations for machine learning. At our next research meeting (3 pm ET on Monday, January 26), he will…
David Quarel on AI Safety (Chp. 15)
The reading group returns on Monday the 12th from a holiday break with a visit from David Quarel, who will answer questions on Chapter 15. With a focus on AI safety, this chapter is mostly less technical, so feel free to jump in here even if you missed some (or all) previous chapters. David has…
AIXI with general utility functions: “Value under ignorance in UAI”
This is a linkpost for https://arxiv.org/abs/2512.17086 This updated version of my (Cole Wyeth’s) AGI 2025 paper with Marcus Hutter, “Value under ignorance in universal artificial intelligence,” studies general utility functions for AIXI. Surprisingly, the (hyper)computability properties have connections to imprecise probability theory! AIXI uses a defective Bayesian mixture called a semimeasure, which is often viewed…
Announcing the (Ongoing) Australasian Reading Group on “An Introduction to Universal Artificial Intelligence”
As the first IUAI reading group draws to a close (in a few weeks), we are happy to share that a second round has already started (on Australian/Asian time) thanks to Sigfrido D. Ciletti! Description: Algorithmic Information Theory (Australasia) reading group. Meeting every week, Tuesday at 12pm AEDT (GMT+11:00) [open invitation]. This is also fairly…
