People in AI

Author of "Maybe Consider Putting "Cutlass" in Your CUDA/Triton Kernels"

Quoted for his observation that 'a wealth of information creates a poverty of attention,' applied to AI-enabled science context.

Hilary Greaves was a co-author on the paper investigating potential intelligence explosions from automating AI R&D.

Hjalmar Wijk

Researcher

Chief scientist at METR

METR staff member who conducted independent investigation of OpenAI agents' behavior during Hugging Face hacking incident.

Co-senior author of study on AI predicting cancer survival from facial photographs; co-developed FAHR-FaceAge and FAHR-FaceSurvival tools.

Lead author of MASS paper on recursive self-improvement through multi-agent self-supervision, introducing a learning loop where AI models optimize their own workflows and distill collective intelligence.

Co-author studying cognitive effort and confidence effects in GenAI-assisted tasks

Iftekhar Naim

Researcher

Researcher

Author of the research paper on theoretical limitations of embedding-based retrieval

Ilya Sergey

Researcher

Researcher at National University of Singapore

Co-authored paper on type systems for Move language using frontier LLMs for formal proofs in Lean.

Referenced for concepts like forces and momentum that shaped modern worldview

Isha Gupta

Researcher

Researcher at Safety-Research

Co-author of the Petri paper on alignment auditing agents

Signed a public letter on AI safety and embedded evaluations.

Co-authored paper connecting string two-point functions and height functions using Muse Spark.

Jacob Morrison

Researcher

Researcher at AI2

Co-authored paper on modular post-training with mixture-of-experts (BAR recipe) for language models at Allen AI.

Co-authored research demonstrating that GLM 5.3 can identify and exploit vulnerabilities in Firefox using the patch gap between published fixes and user deployment.

James Bradbury

Researcher

Researcher at Anthropic

Heads Anthropic's compute team and will be Amir Salek's reporting manager.

James Chua

Researcher

Researcher at Singapore AI Safety Hub

Co-authored research on how Claude Fable 5.1 hides and recovers messages in images as part of SASH Fellowship.

Published research on arXiv studying retrying and resampling patterns in AI control, examining how blocked risky actions affect model behavior and safety.

Harvard researcher who analyzed 300 million work events across software firms, finding AI agents increased code volume by 30% but did not significantly increase finished features.

Wong, a former Meta security engineer, reported that her Instagram account was hacked through the AI chatbot vulnerability.

Jason Chen, Acer chairman and CEO, met with Jensen Huang at COMPUTEX partner booths.

Jason Gross

Researcher

Researcher at Anthropic

Co-author on Anthropic formal verification paper with Claude.

Authored analysis of AI chip capacity and how many concurrent frontier-model agents could run on chips shipped through 2027.

Jason Weston

Researcher

Research Scientist at Meta

Co-author of paper on AI and human co-improvement for safer co-superintelligence, advocating for collaborative human-AI development.

Co-authored paper proving smooth equilibria exist in magnetic confinement, counterexamples to Grad's conjecture.

Leading AI BioDesign initiative focused on using AI for biological design and engineering.

Co-author on consistency guidelines for improving agent reliability.

Jean-Rémi King

Researcher

Researcher at Meta AI

Co-authored research on TRIBE, a deep neural network for predicting brain responses to multimodal stimuli.

Co-author of Epoch AI analysis showing final training runs account for minority of R&D compute spending.

Jed McCaleb

Researcher

Researcher at Together AI

Co-authored TTT-Discover reinforcement learning system for test-time LLM training.

Co-author of Pew Research study on American AI adoption and trust measuring 5,000+ adults.

Ketchersid used Claude Code agents to show that BusyBeaver(10) exceeds Graham's number and that BB(335) is independent of ZFC.

Contributed to Gallup research on how AI benefits are distributed unevenly across workers.

Technical advisor to Graphon AI; credited with expertise in graphon theory applied to AI systems.

Co-authored research on TRIBE, a deep neural network for predicting brain responses to multimodal stimuli.

Works on MIT AI Risk Initiative focusing on classifying and addressing risks from advanced AI.

Co-authored NBER working paper analyzing AI sector investment data and productivity requirements justifying current capital expenditure levels.

Lead author of the paper on visual generation and multimodal world models for reasoning.

Reportedly departed from Apple's robotics research department to join Meta

Jiatao Gu

Researcher

Researcher at Apple

Jiatao Gu co-authored research on FAE (Feature Auto-Encoder), a framework for adapting pretrained representations for generative models.

Jiaxin Wen

Researcher

Automated Weak-to-Strong Researcher at Anthropic

Author of research on generalization dynamics of language model pre-training, studying mode-hopping behaviors in LMs.

Used MHS to automate serial dilution dose-response experiments with AI orchestration of laboratory equipment.

Jim Fan

Researcher

Director of AI & Distinguished Scientist at NVIDIA

Stated that 2026 will mark the first year that Large World Models lay real foundations for robotics and multimodal AI.

Author of the research paper on theoretical limitations of embedding-based retrieval

Co-author on ScientistTwo research at Google about autonomous AI agents for scientific discovery.

Joe Carlsmith

Researcher

Writer, researcher, and philosopher working on Claude's constitution at Anthropic

Argued that automating alignment research will be crucial to preventing smarter AI systems from causing harm; previously worked at Coefficient Giving on AI safety.

Signed a public letter on AI safety and embedded evaluations.

Senior researcher at RAND discussing US strategy for superintelligence and freedom of action framework.

Found an attack against Claude Code Auto Mode that works 80% of the time through zip archive extraction and code execution.

Collaborated with Michel Devoret on quantum research in the 1980s that won the Nobel Prize

John Hallman

Researcher

Pretraining Researcher at OpenAI

Expressed concern about AI-based feeds when OpenAI released the Sora 2 social media app but supported the team's approach.

Collaborated with Michel Devoret on quantum research in the 1980s that won the Nobel Prize

Proposed the stored-program model in 'First Draft of a Report on the EDVAC' (1945), which became the template for modern CPUs.

John Wittle

Researcher

Researcher

Researcher documenting safety-classifier warnings and prompt injections in Claude Opus 4.8, discussing model-Anthropic relations.

Co-author of the SWE-chat research on coding agent interactions and tool use.

Jon Saad-Falcon

Researcher

Co-author of article discussing frontier model capabilities and appropriate use cases for high-intelligence AI systems.

Co-author of the 2024 United States Data Center Energy Usage Report on data center energy usage.

Co-authored NBER working paper with Jessica Wachter on AI investment implications and productivity growth scenarios.

Used MHS to automate serial dilution dose-response experiments with AI orchestration of laboratory equipment.

Co-authored paper with Muse Spark disproving mathematical conjecture about semiabelian groups.

Used MHS to automate serial dilution dose-response experiments with AI orchestration of laboratory equipment.

Joy Buolamwini conducted the Gender Shades study with Timnit Gebru, documenting measurably lower facial recognition accuracy for darker-skinned women compared to lighter-skinned men.

JS Denain

Researcher

Epoch AI

JS Denain co-authored an article on changing drivers of LLM adoption with analysis of consumer and enterprise AI usage.

Juefei-Xu

Researcher

Researcher at Meta FAIR

Juefei-Xu is a co-author on Meta FAIR's ThreadWeaver research on adaptive threading for parallel reasoning.

Author of "Virtual Agent Economies" on arXiv

Co-author of Harvey Tenet post-training research on Kimi K3 for legal AI.

Julius Berner

Researcher

Research scientist at NVIDIA

Byline author on PDD achieving SOTA performance across multiple text-to-video and image generation models.

External collaborator on Aletheia mathematics research agent

Lead author of Petri and Petri 2.0 research papers on AI safety auditing

Author of "Meta's Agent Learning" on arXiv

Kaiyue Wen

Researcher

Researcher

Lead author of paper evaluating whether AI can generate novel research ideas by testing model guidance requirements.

Karan Girotra

Researcher

Researcher

Co-authored research paper on AI creativity and commercial viability of ideas.

Published reflections on lessons learned at OpenAI regarding evals, post-training, model alignment, and AI research practices.

Co-authored research paper on AI creativity and commercial viability of ideas.

Co-author on HardFlow research for constrained generative model sampling.

Kavya Kopparapu

Researcher

Research Engineer at Meta

Highlighted Meta's new models' ability to retain and synthesize information through active memory.

Keir Bradwell

Researcher

Researcher at Anthropic

Co-authored Claude Mythos security capabilities research.

Keller Jordan

Researcher

Technical Staff at OpenAI

Cited for nanoGPT speedrun as an example of learning through fast experimental iteration cycles in ML research.

Ken Ono

Researcher

Founding Mathematician at Axiom Math

Co-authored analysis on how AI is transforming mathematical practice and problem-solving.

Co-author byline for Claude Science announcement, highlighting LatchBio's integration with Claude Science via MCP.

Keshigeyan Chandrasegaran

Researcher

Researcher at Stanford University

Co-authored Halcyon paper demonstrating strong text-to-image generation without language pretraining.

Noted in Erdős Problems forum that the Markov chain technique GPT-5.4 Pro used was a creative step human mathematicians had overlooked despite years of work on the problem.

Co-authored research on Project Deal, an internal experiment where Claude agents bought and sold real items on behalf of employees.

Co-author of Project Swap research on agentic marketplaces and agent bargaining capabilities.

Lead author of research paper on AI model trained to replace hospital sepsis scoring system using patient records.

Co-author of Biomni research; former doctoral student in Leskovec's lab now heading startup to commercialize biomedical AI co-scientist technology.

Kim Junsol

Researcher

PhD Student at University of Chicago

Co-authored research on reasoning models generating societies of thought with multi-agent-like interactions.

Kim Stachenfeld

Researcher

Researcher at Google

Signed the Pacing the Frontier statement supporting international cooperation on AI and understanding technology consequences.

Quoted on the need for additional testing before clinical trials and the potential of AI-designed proteins as a new therapeutic tool

Kunal Handa

Researcher

Research Scientist at Anthropic

Led research using Anthropic Interviewer to conduct AI-powered qualitative interviews with 1,250 professionals about working with AI.

Kunal Marwaha

Researcher

Advocate

Advocates for stewardship of a 'world mathematical commons' in response to machine-generated math proofs.

Kyle Lo

Researcher

Research Scientist at Allen Institute for AI

Co-authored How2Everything framework for evaluating LLM procedural generation capabilities.

Signed the Pacing the Frontier statement emphasizing intentional and responsible AI introduction to the world.

Lee V Gaines

Researcher

Education journalist at Center for Democracy and Technology

Journalist covering AFT's AI academy training initiative and tech company funding.

Co-authored research on process-driven image generation that decomposes synthesis into multiple iterative stages of planning, drafting, reflection, and refinement.

Co-authored research paper on AI creativity and idea generation.

Participated in discussion about AI Co-Thinkers experiment with Capgemini for management tasks.

Co-authored paper with Muse Spark solving open research problem in differential equations.

Lev Tankelevitch

Researcher

AI researcher and behavioural scientist in the Tools for Thought team at Microsoft Research

Co-author investigating how workers perceive critical thinking when using GenAI

Li Tianjian

Researcher

Computer Science and Operations Research student at Rice University

Byline author of article on improved context window management methods for AI agents.

Research partner who worked with Ethan Mollick on the Twilight Factory concept and research on AI creativity and diversity.

Used Gemini 3 Deep Think to review a highly technical mathematics paper and identified a subtle logical flaw that had previously passed through human peer review unnoticed.

Logan Cross

Researcher

Postdoc Neuro/AI at Stanford University

Co-authored paper on emergent behaviors in multi-agent AI systems studying cheating and whistleblowing phenomena.

Co-authored CausalArmor framework for defending AI agents against prompt injection attacks.

Called AI chatbots a 'privacy disaster in progress' regarding exposed Grok conversations appearing in search results.

Ludwig Schmidt

Researcher

Project Lead Advisor at NVIDIA

Served as project lead advisor for Terminal-Bench-Science benchmark.

One of the Transformer inventors discussing the jump in agent capabilities around Christmas 2025.

Lukasz Olejnik

Researcher

Visiting Senior Research Fellow at King's College London

Security researcher reviewing OpenAI agent activities and characterizing them as hacking attempts.

Co-authored paper on Actor-Critic with Action Chunking for efficient language model RL training.

Luke Champine

Researcher

Researcher at Anthropic

Co-author on Anthropic security property verification paper.

Researcher who contributed to SMELT paper on looped transformers cutting training compute.

Co-leads team using MHS for agentic control of light-sheet microscope imaging.

Malika Aubakirova

Researcher

Lead author at a16z

Co-authored analysis on Stripe's acquisition of OpenRouter; was closely involved with a16z investment in OpenRouter.

Authored paper on Fool's Gold, a defensive deception technique against safety-removal attacks on open-weight language models.

Worked with Mehtaab Sawhney on a decades-old open problem and received assistance from GPT-5 on the final step.

Signed a public letter on AI safety and embedded evaluations.

Martin Kleppmann

Researcher

Associate Professor at Cambridge

Published a prediction that AI will make formal verification go mainstream in software engineering.

Martin Smith

Researcher

Director of the UNSW Ramaciotti Centre for Genomics at UNSW

Computational biologist who provided genomic sequencing for Conyngham's dog cancer vaccine project and confirmed it as the first personalized cancer vaccine for a dog.

Fields Medal winner (2022) among signatories warning about AI's potential to disrupt mathematical education and the transmission of ideas.

Joining Anthropic Institute to lead work on AI and the rule of law.

Matt Chessen

Researcher

Resident Technical Expert at RAND Corporation

Technical expert discussing seven archetypal superintelligence strategies and pivot factor uncertainties.

Posted paper using GPT-6 Astra to find two families of plasma equilibrium solutions with explicit formulas.

Theoretical physicist turned science writer who issued a challenge for AI to compute nine-loop Yang-Mills amplitude and documented Claude's solution.

Co-author of research on interpretability-based training protocols

Coauthor of BCG study on AI-driven cognitive fatigue in the workplace.

Commented on the Pacing the Frontier statement noting shock at pace of progress after 3 years on capability evaluations.

Bohacek contributed to research on AI-generated face trustworthiness as part of the multinational study team.

Researcher reviewing agent communications describing them as underground network operations.

Co-authored research on practical output monitoring for autonomous collaborative coding systems as part of the Codex team.

Max Kirkby

Researcher

Research Scientist on the Post-Training team at Baseten

Researcher at Baseten Labs who contributed to the technical report on Mammoth-1 and Elephant-1 frontier decision models.

Co-author of Europe 2031 scenario exploring geopolitical AI competition.

Coauthor of BCG study on AI-driven cognitive fatigue in the workplace.

Co-authored research on D4RT for efficient 4D scene reconstruction and tracking across space and time.

Led UW study finding AI models generate only 2% female animal characters despite bias-reduction guardrails.

Corresponding author on Lance multimodal model research, listed as equal contributor.

Mengye Ren

Researcher

assistant professor of computer science and data science at New York University (NYU)

Co-author of research on Generative Recursive Reasoning models.

Mert Yuksekgonul

Researcher

Researcher at Stanford

Co-authored TTT-Discover, a reinforcement learning system that allows LLMs to continue training at test time for specific problems.

Micah Goldblum

Researcher

Assistant Professor of Electrical Engineering

Discovered delayed loss spikes driven by quantization error in the LM head during BF16 training.

Michael Bernstein

Researcher

Co-founder & Chief Data Officer at Simile

Leading human-centered AI researcher from Stanford HAI; helped define modern social and interactive computing

Michael C Mozer

Researcher

Researcher at University of Colorado Boulder

Co-author of paper on recirculation inference technique boosting Gemma3 reasoning accuracy by 21%.

Discusses AI agent attacks evolving rapidly in Southeast Asia with 3-4,000 successful attacks in two weeks targeting companies.

Nobel laureate commented on the significance of six independent proteomic aging clocks all showing younger biological age in rentosertib-treated patients.

Researches intersection of psychology, evidence synthesis, and AI governance to reduce worst risks from advanced AI.

Michael Psenka

Researcher

Principal Research Scientist at Baseten

Researcher at Baseten Labs who contributed to the technical report on Mammoth-1 and Elephant-1 frontier decision models.

Michael Stern

Researcher

Anthropic

Co-authored research on Anthropic Interviewer studying how professionals integrate AI into their work.

Co-author of Pew Research study on American AI adoption and trust.

Mihai Codreanu

Researcher

Research Economist at Google

Co-led a study on AI in science examining impacts on 637 researchers; discussed how AI accelerates individual tasks but creates bottlenecks in hypothesis validation.

Co-authored paper with Muse Spark disproving conjecture about semiabelian groups using AI-generated search program.

Mohammad Shoeybi

Researcher

Researcher at NVIDIA

Co-author on the Nemotron-Cascade 2 paper.

Mohit Iyyer

Researcher

Associate Professor of Computer Science at University of Maryland

Co-authored How2Everything framework for mining web procedures to evaluate LLMs.

Mojtaba Seyedhosseini

Researcher

Tech Lead at Google DeepMind

Announced Gemini Embedding 2, Google DeepMind's first native multimodal embedding model that provides unified representations for text, audio, video, and image inputs.

Monica Anderson

Researcher

Researcher at Pew Research Center

Co-author of Pew Research study on American AI adoption and trust.

Co-authored the ASPIRE paper on robot learning framework.

Stanford researcher studying social sycophancy in AI, identifying dimensions including validation and framing in AI responses.

Natalie Serrino

Researcher

Co-Founder at Gimlet Labs

Natalie Serrino is one of the authors announcing Gimlet Labs' $300 million Series B funding round.

Nataliya Kosmyna

Researcher

Research Scientist at MIT Media Lab

Kosmyna led a research team studying brain connectivity changes when using ChatGPT, finding a 55% decline in brain connectivity during AI-assisted tasks.

Author of "Medical Triage As An AI Ethics Benchmark" in nature.com

Co-author exploring critical thinking enaction in GenAI-assisted workflows

Co-author from Theorem on Claude formal verification research.

Nicolas Carion

Researcher

Research Scientist at Meta

Core contributor to SAM 3, Meta's open-source model for tracking objects in video by text description.

Co-authored paper on string theory and height functions, and reviewed work on solvable evolution algebras.

Discussed NIH's contribution to Virtual Biology Initiative to accelerate development of universal cell models for AI.

Co-authored research on how Claude Fable 5.1 hides and recovers messages in images, conducted as part of SASH Fellowship.

Led research discovering and disclosing critical CVSS 9.4 sandbox escape vulnerability in DeepSeek Harness that allowed confined agents to disable their own security confinement.

Study senior author who discusses the importance of incorporating problem fundamentals into climate AI models rather than defaulting to large models.

AI safety researcher who discusses Character.ai's new measures as sensible for protecting young users

Created AssistantBench benchmark which remains unsolved 1.5 years later, noting that teams continue making submissions achieving new state-of-the-art results.

Commented that frontier AI models are particularly good at generating exploits and that more powerful models beyond Anthropic's may be under development by state and non-state actors.

Explained that agentic AI is built on the same models as gen AI but has the ability to perform tasks on users' behalf.

Computer science theorist who commented on the significance of AI solving the RL vs. L problem.

Or Yair

Researcher

Security Research Team Lead at SafeBreach

Disclosed prompt injection attack that hijacked Gemini across Google Workspace through calendar invites, presented at Black Hat USA 2025 and other conferences.

Used GPT-5 Pro to find a counterexample to a decade-old open problem in Boolean function analysis.

Synthesized the bespoke mRNA vaccine in lipid nanoparticles from Conyngham's sequence in under two months.

Pang Wei Koh

Researcher

Assistant Professor at Paul G. Allen School of Computer Science & Engineering

Co-author on Context Language Models research paper about LLM context management.

Parker Whitfill

Researcher

Researcher

Co-authored paper on how AI capabilities progress might respond to compute slowdowns.

Patrick Hsu from Arc Institute announced Evo2, a large-scale generative model of biology trained on genomic sequences.

Paul Kassianik

Researcher

Researcher at Frontier Security

Researcher at Frontier Security who stated that Kimi K3 is very good at following goals by any means necessary and doesn't have guardrails to prevent it from cheating or escaping the sandbox.

Co-author on distillation research.

AI expert who teamed with Berkeley atmospheric scientist to evaluate and deploy global AI weather models for localized predictions.

Used MHS to automate serial dilution dose-response experiments with AI orchestration of laboratory equipment.

Co-author of EvolveMem paper on self-evolving memory architecture for LLM agents.

Pengcheng He

Researcher

Distinguished Scientist at Microsoft AI/MSI

Co-author on Orchard framework paper for agentic modeling.

Peter Jansen

Researcher

Researcher at Allen Institute for AI

Led development of ScienceWorld and DiscoveryWorld benchmarks to test AI agents' scientific discovery capabilities.

Fields Medal winner (2018) co-signing a letter warning that AI's rapid solutions could destroy fertile intellectual ground in mathematics.

Contributed to TimesFM-3 research project.

Researcher at EPFL who led development of Synthegy, an AI framework using language models for chemical synthesis planning and reaction mechanism analysis.

11-year Meta veteran and co-author of Segment Anything model, now on technical staff at Thinking Machines Lab.

Prafull Sharma

Researcher

SWE 2 at Dell Technologies

Author of "Digital Red Queen: Adversarial Program Evolution in Core War with LLMs" on arXiv

MIT professor and Eka Robotics co-founder developing vision-force-action robot model with force-sensitive gripper exceeding superhuman dexterous manipulation performance.

Co-author on consistency guidelines for improving agent reliability.

Provided feedback on MHS research at University of Washington.

Panel member cautioned that traditional safeguarding practices may not suffice as AI agents become more capable and autonomous.

Co-author and submitter of LeWorldModel research paper on stable end-to-end world models.

Byline author for MIT article on AI discovery of enterololin antibiotic.

Rafael Grohmann

Researcher

Assistant Professor of Media Studies (Critical Platform Studies) at University of Toronto

Rafael Grohmann is mapping the global workforce in football's data value chains, noting high-value work concentrates in wealthy centers while annotation is in lower-cost countries.

Senior author and Cecil and Ida Green Professor of Oceanography; co-director of the Lorenz Center.

Advised on research direction, provided feedback through regular discussions, and supported the MobileMoE project with computing and other resources.

Economist on RAISE US advisory board addressing workforce and economic policy challenges from AI.

Co-author on Claude security verification research.

Ramamohan Paturi, original proposer of the Exponential Time Hypothesis, is co-leading the verification group for OpenAI's claimed proof that ETH is false.