
Junior Professor @SorbonneParis1 Philosophy
Paris, France
Joined April 2021
- Tweets109
- Following54
- Followers151
- Likes113
A great essay by @Dr_Atoosa!
💥 Thrilled to share my new essay "A New Dawn For Mathematical Progress" out today in @NoemaMag! Read the essay here: noemamag.com/
As frontier AI systems solve some of the most difficult open questions in mathematics such as the Millennium Prize problems, many fear human mathematicians are becoming obsolete.
I argue that whether AI ends human mathematics or begins its golden age depends on which of two models of "mathematical progress" we embrace: the checklist model or the generative encounter model.
Right now, the impacts of AI systems on mathematics has left the mathematical community in deep disorientation. A PhD student recently wrote of a "profound spiritual crisis"; a Fields medalist called for "new theories of care and love."
These reactions make sense under the checklist model, where progress consists of resolving well-defined, pre-specified problems (Olympiads, famous conjectures, Millennium Prizes) to crown lone human heroes.
However, the checklist model was always a narrow view; one that elevated "pure" math as the only "real" math, while treating "applied" math as secondary.
As the mathematician George Dantzig once said: "I consider myself a mathematician. I don’t differentiate between pure mathematics and applied mathematics."
History shows that so much of our richest math comes from the generative encounter model on which mathematics makes breakthroughs when human mathematicians tackle complex and messy real world problems.
If we embrace the generative encounter model of mathematical progress, a golden age of mathematics, rooted in human purpose & new visions of education & collective reward, may have just begun.
1
83
Mario Günther retweeted
💥 Thrilled to share my new essay "A New Dawn For Mathematical Progress" out today in @NoemaMag! Read the essay here: noemamag.com/
As frontier AI systems solve some of the most difficult open questions in mathematics such as the Millennium Prize problems, many fear human mathematicians are becoming obsolete.
I argue that whether AI ends human mathematics or begins its golden age depends on which of two models of "mathematical progress" we embrace: the checklist model or the generative encounter model.
Right now, the impacts of AI systems on mathematics has left the mathematical community in deep disorientation. A PhD student recently wrote of a "profound spiritual crisis"; a Fields medalist called for "new theories of care and love."
These reactions make sense under the checklist model, where progress consists of resolving well-defined, pre-specified problems (Olympiads, famous conjectures, Millennium Prizes) to crown lone human heroes.
However, the checklist model was always a narrow view; one that elevated "pure" math as the only "real" math, while treating "applied" math as secondary.
As the mathematician George Dantzig once said: "I consider myself a mathematician. I don’t differentiate between pure mathematics and applied mathematics."
History shows that so much of our richest math comes from the generative encounter model on which mathematics makes breakthroughs when human mathematicians tackle complex and messy real world problems.
If we embrace the generative encounter model of mathematical progress, a golden age of mathematics, rooted in human purpose & new visions of education & collective reward, may have just begun.
☄️Come work with me in Paris! Two-year postdoc in Philosophy of AI, broadly understood, at Panthéon Sorbonne.
Apply by 17 Dec 2026 for full consideration. I’d love to study your ideas. Apply here:
philjobs.org/job/show/32693
Please share!
Mario Günther retweeted
Who owns the benefits of AI? In this essay I co-authored with @IasonGabriel for the DeepMind Institute, we offer four moral arguments — grounded in rights, reciprocity, fairness, and beneficence — that together make the philosophical case that the benefits of AI should reach everyone.
Read the piece here and let us know what you think: institute.deepmind.com/essay…
Here is the "latest development in decision-theoretic accounts of legal proof" which builds on a notion of retributive justice. #LegalTheory #DecisionTheory #LegalProof #RetributiveJustice
📖 Presentation of the article in Quaestio facti No. 11: «Bayesian decision theory can guide legal factfinding», by @Dr_MarioG. Professor Günther, from Université Paris 1 Panthéon-Sorbonne, presents his contribution published in the latest issue of the journal: revistes.udg.edu/quaestio-fa…
Mario Günther retweeted
OpenAI was founded because Sam and co thought Google couldn't be trusted with AGI. Anthropic split off to be the "responsible" safety-focused alternative which also became "we're building the best AI”. Now the two companies that set out to slow Google down are racing each other. Meanwhile my non-AI friends who work in domains like climate environmentalism or independant film industry have become anti-AI, have lost their imagination of AI helping them to proceed in their goals because they see a parade of "AI researchers say AI will make humanity extict" headlines and conclude that any AI product companies ship must be dangerous. Then after years and literally years of proper discourse with AI safety and EA and rationalist folks that watch your words and don’t confuse the world, some of their loud voices still equate cheaply AI risk with AI extinction risk in their communication and mind; “real risk” is now a catch all for extinction, exsitential, bioterr, bad cyber attacks, takeover of governments and shoot happened because sanbox was exploitable oh and really bad AI risk. This cheap talk and confusion generation approach has been one of their worst disservices to the world. Lesson learnt: it’s gonna be quite hard but stay grounded and correct the public imagination. Wrong fear is one of the worst types of fear. It could guide you down the wrong road and it could take away your capacity for clear thinking and correct acting before any AGI takes that away from you.
Mario Günther retweeted
Must one be an Effective Altruist (EA) or a Yudkowskian Rationalist (YR) to give credibility to AI x-risk claims? Not necessarily. A 🧵
When should you explain AI agents by ascribing them human-like capacities? Here is @Dr_Atoosa and my proposal.
Over the past few days, @Dr_MarioG and I read and learnt quite a lot about anthropomorphisms in AI explanations from many people, in particular @dwarkesh_sp @ajeya_cotra @HeidyKhlaaf @anilkseth @ccatalini. We couldn't fit our thoughts into an X 🧵 so we wrote a blog post: atoosatopia.substack.com/p/a…
Mario Günther retweeted
Everyone says 2026 is the year of AI agents. But what are they, and how should we govern them? @IasonGabriel and I have answers for you in our new Nature paper.
(1) We propose thinking about AI agency through the lens of an "agentic profile," defined by four core constitutive properties: autonomy, efficacy, goal complexity, and generality.
(2) We use these properties to construct distinct profiles for four different types of AI agents.
(3) We outline practical ways to measure agency along these dimensions.
(4) We argue that AI governance cannot rely on a one-size-fits-all approach, demonstrating why it must instead be tailored dynamically to match each agent's unique profile.
Read the full paper here: nature.com/articles/s41586-0…
Reply published: Holger Andreas and I respond to the commentaries on our paper “Factual Difference‑Making” in the Australasian Philosophical Review, refining our analysis of causation to meet the counterexamples raised. Read more: wix.to/T6ZrcLK #philosophy #causation
3
68
I moderated “Protecting Vulnerable and Underserved AI Users” at IASEAI'26. Watch the full session here: wix.to/oM6omXC #IASEAI26 #AIethics #ResponsibleAI
1
43
Presented joint work w/ Sven Neth at UBC — "Equal Say as Sum Normalization." Great feedback & thanks to Pamela Robinson! Read more: wix.to/ELW8iSM #ResearchTalk #UBC #Math
2
95
New open-access paper with Conrad Friedrich — “A Bayesian Justification for the Scenario Approach to Legal Proof.” We offer a Bayesian framing of the scenario approach to strengthen its use in legal proof. Read it here: wix.to/0IISuM2 #LegalTheory #AcademicResearch
1
35
Mario Günther retweeted
The "third story" about AI risk deserves a lot more attention. It doesn't pretend that we can predict with certainty exactly what's coming. It doesn't act as if "ethical" risks are somehow separable from "safety" risks. The world is a complex system! /FIN
vox.com/future-perfect/46168…
My lead article "Factual Difference-Making" is now published in the Australasian Philosophical Review. It proposes a new research program for analysing causation! Read it here: wix.to/ftqHQkY #Philosophy #Research
2
52
Mario Günther retweeted
I was planning to launch my substack on "Human, life, AI, and future" in a few months, with something very different. I’ve been working quietly on some exciting research about AI and the future of humanity—big questions, long arcs, and some surprising ideas I was excited to share when the time felt right. But life doesn’t always care about our timelines. Events erupt. Emotions build. And suddenly, waiting feels like avoidance. So here is my first post: "Valid and sound inference is a form of love: On War, ignored evidence, and the seduction of shallow narratives"
Read here: atoosatopia.substack.com/p/v…
Mario Günther retweeted
Ever since I first heard the slogan “AI as normal technology,” I’ve felt uneasy. Tonight that unease crystallised. In this 🧵I unpack what normal hides and why the metaphor may ultimately fail to capture AI’s abnormal impacts on human life and societies. 1/n
I have the pleasure of commenting on Minkyung Wang's paper "Accuracy and Informativity of Belief" at the Formal Epistemology Workshop in Iceland. Check out more details: wix.to/eV0Khzz #FormalEpistemology #Iceland
1
54
How does the probability of guilt influence legal verdicts? I tackle this crucial question in my latest article published in the Canadian Journal. Check it out: wix.to/unvv5jJ #CriminalJustice #Law #ProbabilityOfGuilt
1
49
What should determine a defendant's guilt? My answer is published in Legal Theory. Read here: wix.to/rqcRJLk #LegalProof #Justice #Law
1
61









