Blog
Articles and reflections on the evolving intersection of artificial intelligence and fundamental science.
Lab Notes: CiteVerify - Fake citation detection and evidence verification for AI research
Scientific writing has always relied on the implicit contract that claims must remain traceable to evidence. Large language models complicate that relationship. Modern systems can generate reports that appear rigorous and extensively sourced while quietly introducing fabricated citations. CiteVerify is FirstPrinciples’ answer to the rise of fake citations.
Automated conjecturing: How machines are exploring mathematical structure
Before a theorem can be proved, someone has to decide what is worth proving. For decades, the formation of a conjecture remained largely outside the reach of machines. In Automated Conjecturing with TxGraffiti, Randy Davila explores how that boundary has shifted.
.png)
Lab Notes: Teaching models to reason a little better
A recent recap of work at FirstPrinciples, from fine-tuning and RLVR to ensembles and 122B-scale models. We’re beginning to clarify what improves scientific reasoning, and these are our notes.
Between tools and theory: Reflections from the Machine Learning and the Physical Sciences workshop, NeurIPS 2025
A reflection on ML4PS 2025, where researchers in physics and machine learning grappled with the role AI should play in scientific discovery, and what it would take to move from process acceleration toward deeper scientific insight.
The AI Physicist reaches its first autonomy milestone (ahead of schedule)
The team at FirstPrinciples has reached the AI Physicist’s first autonomy milestone ahead of schedule, assembling the early loops of a system that will guide its own scientific reasoning from research to hypothesis and beyond.
The case for specialization: Building scientific AI that thinks like a physicist
Large Language Models have changed how we think, work, and do science, but can they truly reason like scientists? At FirstPrinciples, we’re exploring the limits of AI generalization and the promise of specialization through the development of the AI Physicist.
Chain-of-thought seen as key to AI safety, but experts warn it’s fragile
Chain-of-thought reasoning has become a rare interface between human and machine logic. But a new paper warns that the window may be closing.
AI and openness at CERN: FirstPrinciples demos the AI Physicist at the Open Science Fair
FirstPrinciples presented an early demo of its AI Physicist at the Open Science Fair, held this year at CERN. The event sparked conversations on trust, openness, and the role of AI in research, underscoring how collaboration will shape the future of discovery.
AI enters the scientific loop: Simulation, integrity, and the rise of open reasoning
From prompt injection to physics simulators and open reasoning models, recent news shows that AI isn’t just accelerating science, it’s reshaping how it works. The question now, is will it deepen inquiry, or erode the principles on which credibility in science is built?
The physics of AI hallucination: New research reveals the tipping point for large language models
Physicist Neil Johnson has mapped the exact moment AI can flip from accurate to false, and he says understanding the physics could be the key to safer systems.
New AI models and the benchmark paradox
Z.ai's new open-source model and Harmonic's math-native chatbot highlight contrasting strategies for AI reasoning, while a new wave of increasingly specialized benchmarks invites us to rethink how we measure progress.
AI faces a tough physics exam: New benchmark reveals the challenge
Large language models have advanced dramatically in recent years, yet when physicists gave them an undergraduate-level test, even the best models were only correct on around one out of every three questions. The new PhysUniBench benchmark exposes how far AI still has to go in mastering fundamental science.
Prediction isn’t understanding: AI’s evolution and the soul of science
From rule-based ‘expert systems’ to neural networks, AI has long chased the dream of scientific reasoning. But while today’s models are good at pattern matching and can generate code or summarize academic papers, they struggle with the heart of scientific discovery: structured reasoning.