Formal Certification Technologies for AI Safety

Developing the software infrastructure to make AI systems safe, with formal guarantees

Fund thebes' model tinkering

Research, tutorial writing, and open-source libraries & tools for experimenting with language models

FloraForge: Aesthetic Gene Editing in Petunias

A modular gene-editing platform for engineering new pigment patterns in ornamental plants, starting with a vein-pattern rescue line in petunias.

6 Month Stipend to Support a Transition to AI Governance Work

Funding research on how AI hype cycles can drive unsafe AI development

Exploring a Single-FPS Stability Constraint in LLMs (ZTGI-Pro v3.3)

Early-stage work on a small internal-control layer that tracks instability in LLM reasoning and switches between SAFE / WARN / BREAK modes.

Multilingual Open Textbooks

SDCPNs for AI Safety

Developing correct-by-construction world models for verification of frontier AI

Software to detect data fabrication

Continue developing ‘copy-paste-detective’ - a software that detects signs of data fabrication in scientific research - and run it against 20,000 publicly available Excel datasets.

Disentangling Political Bias from Epistemic Integrity in AI Systems

An Integrative Framework for Auditing Political Preferences and Truth-Seeking in AI Systems

Eradicate the hidden factory farming of feeder rodents

Developing and commercializing an alternative to feeder rodents—tackling one of the most neglected and high-suffering forms of factory farming.

OpenReg: AI tools for FDA application and reg compliance

A set of AI-powered open-source tools intended to shorten time-to-market for medtech innovators.

Charting the Policy Landscape of Advanced Education in America

This nationwide report will provide a detailed account of the policies states have that in place to help advanced learners thrive.

Search for more solar system objects

Purchase compute and support algorithm development

Global Governance & Safety Layer for Advanced AI Systems - We Stop Rogue AI

Building the first external oversight and containment framework + high-rigor attack/defense benchmarks to reduce catastrophic AI risk.

Clean Indoor Air for Schools

Accelerating the adoption of air filters in every classroom

When Safety Testing Costs More Than Model Training

Multi-agent AI security testing that reduces evaluation costs by 10-20x without sacrificing detection quality

Fund a Fellow for the Cooperative AI Research Fellowship!

A 3-month fellowship in Cape Town, connecting a global cohort of talent to top mentors at MIT, Oxford, CMU, and Google DeepMind

Finishing a Video on the Google DeepMind Hunger Strike

Funding gap to pay for a video editor and scriptwriter

Enabling a Student to Present his Mechanistic Interpretability Work at NeurIPS

Help fund our student’s trip to NeurIPS to present his main conference paper on interpretable features in text-to-image diffusion models.

Synthesizing Standalone World-Models

Research agenda aimed at developing methods for constructing powerful, easily interpretable world-models.