An AI Safety Expert Just Shared 4 Years of Research — Free For Anyone

An AI safety researcher just published a public reading list after spending four years studying how artificial intelligence might go wrong — and what we can do about it. The list includes research papers, YouTube channels, podcasts like the 80,000 Hours Podcast, free online courses, and statements from major organizations like the Center for AI Safety and Anthropic. It's essentially a beginner's roadmap to a topic that's usually locked behind academic jargon and paywalls.

The list is broken into categories: technical papers explaining how AI systems can develop dangerous goals, videos that make the concepts visual and accessible, podcasts featuring researchers in conversation, and even petitions everyday people can sign. Some items are in French, and the author admits not everything is peer-reviewed — this is one person's curated favorites, not a formal curriculum. Still, it's a surprisingly approachable entry point.

Why should this matter to you? AI safety isn't just a Silicon Valley obsession anymore. Governments are drafting AI laws, companies are racing to build more powerful systems, and experts — including some who built these tools — are publicly warning about risks ranging from job disruption to loss of human control. Understanding the language of these debates helps you follow the news, vote on policy, and make smarter decisions about which AI tools to trust with your data and your work.

The catch: this list is dense. It's not a fun weekend read — it's a syllabus. Treat it like a library, not a checklist. Start with one YouTube video or podcast episode, and see where curiosity takes you. If AI headlines confuse you, this is your chance to get ahead of the curve without quitting your day job.