Pat Pataranutaporn

I'm excited to share that I have been appointed as a new MIT faculty starting this September as an assistant professor at the MIT Media Lab. Very excited! Official Announcement: media.mit.edu/posts/patpat-p… @medialab @FluidInterfaces @AHA_MediaLab @MIT

We’ve been quietly working on projects we weren’t ready to share until now. The event is by invitation only. A public livestream will be available for those who want a glimpse of what’s coming: media.mit.edu/events/aha-sym…

Human Capital in the Age of AI Asia Business Speaker Series 2026 Continues by exploring AI and Tech Space! 📍 Location: Aldrich 008, Harvard Business School 📅 Date: Tuesday, April 21 ⏰ Time: 3:30 – 4:00 PM EDT 🎟️ Secure your spot: luma.com/s5aj3ule

OMG!!! We concluded with: “At present, only 16.0% of reviewed benchmarks conducted any statistical testing. Increasing the use of robust statistical methods for LLM benchmarking is critical.” @GaryMarcus

@elonmusk — Frank Herbert, Dune Messiah (1969) He knew.

The AI Scientist: Towards Fully Automated AI Research, Now Published in Nature Nature: nature.com/articles/s4158… Blog: sakana.ai/ai-scientist-n… When we first introduced The AI Scientist, we shared an ambitious vision of an agent powered by foundation models capable of executing the entire machine learning research lifecycle. From inventing ideas and writing code to executing experiments and drafting the manuscript, the system demonstrated that end-to-end automation of the scientific process is possible. Soon after, we shared a historic update: the improved AI Scientist-v2 produced the first fully AI-generated paper to pass a rigorous human peer-review process. Today, we are happy to announce that “The AI Scientist: Towards Fully Automated AI Research,” our paper describing all of this work, along with fresh new insights, has been published in @Nature! This Nature publication consolidates these milestones and details the underlying foundation model orchestration. It also introduces our Automated Reviewer, which matches human review judgments and actually exceeds standard inter-human agreement. Crucially, by using this reviewer to grade papers generated by different foundation models, we discovered a clear scaling law of science. As the underlying foundation models improve, the quality of the generated scientific papers increases correspondingly. This implies that as compute costs decrease and model capabilities continue to exponentially increase, future versions of The AI Scientist will be substantially more capable. Building upon our previous open-source releases (github.com/SakanaAI/AI-Sc…), this open-access Nature publication comprehensively details our system's architecture, outlines several new scaling results, and discusses the promise and challenges of AI-generated science. This substantial milestone is the result of a close and fruitful collaboration between researchers at Sakana AI, the University of British Columbia (UBC) and the Vector Institute, and the University of Oxford. Congrats to the team! @_chris_lu_ @cong_ml @RobertTLange @_yutaroyamada @shengranhu @j_foerst @hardmaru @jeffclune

The line-up for the AI and Innovation series at the Ryan Institute on Complexity in the Spring is pretty epic! Mark your calendar for talks by @zhitzig, @allisonkoe, @samgarvingold, @patpat_mit

I am honored to join the panel at Cornell's Health Tech Summit, "Moonshots: Big Bets and Radical Innovations Transforming Healthcare," alongside Jianying Hu (IBM), Justine Levin-Allerhand (Flagship), and Guillermo Sapiro (Apple): healthtechsummit.ai/2026-program @medialab @AHA_MediaLab

I am deeply honored to be speaking today at the UN Economic and Social Council (ECOSOC) Chamber on “AI and the Future of Human Flourishing,” as part of the Royal Academy of Science International Trust (RASIT) event #UnitedNations

My first publication in Nature! I learned so much contributing to this one...

Thank you Irmandy Wicaksono for hosting me tomorrow at National University of Singapore. Please come if you are around!

If you're in Singapore tonight, join me and Kathleen Ditzig (National Gallery Singapore) for this conversation on AI, Art, and human flourishing. 📍 Padimai Art & Tech Studio, 39 Keppel Road #03-02, Singapore 089065 📅 31 January 2026, 5–7pm

Visiting Shenzhen and speaking at Scalable HCI at SUSTech School of Design was amazing. It felt like stepping into the future. scalablehci.com/2026/program/

🚀 CHI 2026 BiAlign Workshop & Happy Holidays🎄🥰! Thrilled to announce the #CHI2026 BiAlign Workshop on Human–AI Interaction Alignment! We invite researchers & practitioners to join us for Designing, Evaluating, and Evolving Value-Centered AI for Reciprocal Human–AI Futures.✨ 📅 Deadline: Feb 12, 2026 (11:59pm AOE) 📄 Submission formats: 2-page (tiny), 4-page (short), & 8-page (long) papers 📍 Workshop date: Mid-Apr, 2026 (TBD) | CHI 2026 🔗 Website: bialign-workshop.github.io We welcome work on: • Value-centered alignment mechanisms • Novel alignment demands for responsible AI • Design and interaction for alignment; • Evaluating alignment and societal impacts; • Dynamic co-evolution of human-AI futures 👫 Join an interdisciplinary community spanning HCI, AI, ML, NLP, CSCW, XAI, ethics, and social computing, and help shape the future of reciprocal, value-centered human–AI alignment. 💗Many thanks to —and meet—our amazing organizing team: @tknearem, @divy93t, @patpat_mit, @Yike_Shi, @jennytliang, @Anoop_Sinha, @tanmit, @_lamaahmad, @bradamyers, @yangli169! #CHI2026 #BiAlign #HumanAIAlignment #ValueCenteredAI #CFP #HCI #ResponsibleAI

Your 3am questions were... interesting. Your 2025 AI Wrapped is ready and we're not judging (much) 😏 aiwrapped.org #aiwrapped2025 Privacy note: Your conversation data is de-identified before analysis and is not stored. The analysis is for research purposes only.

Tomorrow, I'll be sharing my most recent research from the MIT Media Lab at the University of Tokyo, thanks to Professor @drinami for the hosting me there. Time: 12:00 PM Location: University of Tokyo, Living Lab Komaba

I'm thrilled to share that I'm in Tokyo this week to receive the AI Creative Future Grand Prix Award! Learn more about the award: aica-awards.com Explore the project: cybersubin.media.mit.edu

🚨New Preprint 🚨 Current agent evals mostly measure competence, but miss behavior e.g. are their decisions stable, rational, manipulable, human-like? We introduce ABxLAB, a framework for studying agent behavior. Using it we create an agentic consumer behavior benchmark. 🧵1/9

Introducing NeuroChat: the world’s first neuroadaptive chatbot that adapts its responses to your cognitive engagement. By reading brain signals in real time, NeuroChat personalizes its teaching style to your attention, curiosity, and focus. NeuroChat works with your brain, not around it. Developed at the MIT Media Lab.

The performance of generative A.I. models for clinical reasoning are not holding up to increased scrutiny arxiv.org/abs/2509.18234 @MSFTResearch ai.nejm.org/doi/full/10.10… @NEJM_AI @AdamRodmanMD @LiamGMcCoy

At this year's @TheAtlantic Festival, @HannaRosin and I explored how technology shapes Human Flourishing in both profound and problematic ways. Full conversation: youtube.com/watch?v=b128U0… media.mit.edu/posts/the-atla… #TheAtlantic #MITMediaLab