YouTube
A collection of essential videos covering AI safety, ethical considerations, and the societal impact of artificial intelligence.
Geoffrey Hinton on AI Dangers
A seminal discussion on the existential risks posed by advanced artificial intelligence. (Source: The Diary of a CEO)
Mo Gawdat on AI Dystopia and Utopia
An outline of potential dystopian and utopian futures shaped by AI. (Source: The Diary of a CEO)
Dr. Roman Yampolskiy on AI Safety
An in-depth analysis of superintelligence risks and projected AI safety timelines. (Source: The Diary of a CEO)
What OpenAI Doesn’t Want You to Know
An investigative report on the ethical controversies associated with OpenAI. (Source: More Perfect Union)
The Chinese Room Is a Dishonest Argument
A philosophical examination of the Chinese Room Argument and artificial consciousness. (Source: Curt Jaimungal)
The Dark Side of AI Data Centers
An exposé on the environmental and societal impact of AI data infrastructure. (Source: Business Insider)
Courses
Free educational courses to build expertise in AI alignment and safety, based on 2025 recommendations.
AI Alignment (BlueDot Impact)
A foundational course on core AI safety concepts, evaluation, and structured debate.
AI For Everyone (Coursera)
A non-technical overview of AI capabilities, limitations, and ethical considerations.
Elements of AI (Univ. of Helsinki)
A broad introduction to AI, including key principles of ethics and alignment for a general audience.
Intro to ML Safety
A technical introduction to modern machine learning safety and alignment techniques.
AI Safety Fundamentals
A comprehensive curriculum covering the AI alignment problem in depth.
Datasets & Reports
Open resources for AI safety research, including benchmarks and risk assessments from 2025.
2025 AI Index Report (Stanford HAI)
A comprehensive annual report on trends, risks, and progress in artificial intelligence.
2025 AI Safety Index (Future of Life)
An assessment of the safety and transparency efforts of leading AI development companies.
AI Safety Institute (UK) Reports
Official government research on frontier AI risks and model safety evaluations.
Anthropic's HH-RLHF Datasets
Datasets for training models to be helpful and harmless using reinforcement learning from human feedback.
OpenAI Safety Reports
Official documentation on safety evaluations and red teaming for models like o1.
AI Safety Center Benchmarks
A suite of benchmarks for evaluating risks such as self-replication and goal misinterpretation in LLMs.
Research
Details about our ongoing research projects and publications will be available here soon.
Meet the Team
The dedicated team driving AI safety innovation at Labonsky AI Research.
Marcin Labonski
Founder & CEO
Directing the organization's research strategy in AI alignment and pioneering novel approaches to risk mitigation.
Meet the Research Assistants
Our highly-valued team members, providing moral support and expert-level napping.
Lia
Chief Morale Officer
Levi
Lead Sleep Analyst
Lutka
Head of Box Fort Architecture
Rengar
Junior Pounce Engineer
Lilith
Feline Language Model