AI safety has a hiring efficiency crisis: public work outpaces résumé submissions
Because the talent market is inefficient and hard to screen, public writing, projects, and networking get you in front of AI safety recruiters faster than mass applications.
The video won't play here. Listen to the audio instead:
The argument · timestamps estimated from transcript position
Having an informed viewpoint matters more than becoming an expert first
Building foundational AI safety literacy matters more than deep expertise. The target: hold a 30-minute conversation at an AI safety conference and convince someone you understand the core problems and a few solution paths. You need your own views on AI's progress speed, the most urgent problems, and which institutional theories make sense—you don't need certainty, but you need a reasoned position. Applicants who only repeat correct answers without showing their thinking tend to get screened out early.
— Matt BeardSmall cheap experiments reveal your fit better than career planning
The article images three precarious ladders leaning on each other, captioned: ‘It's fine if your career looks unstable for a while.’Instead of debating which path suits you—technical, policy, or communications—test with weekend experiments: write a blog post analyzing a think tank report, build a reusable data dashboard, organize a local meetup. Boredom or exhaustion from an experiment is signal too. Expand your options first; experiments and outside feedback will narrow them down.
— Matt BeardThe field grew 30× in 8 years; you're probably undershooting your chances
AI safety has more incubators, fellowships, and job listings than any other domain. The ecosystem grew approximately 3000% in 8 years, powered by problem urgency and a sprawling talent-development infrastructure. Yet the author, who advises in the field, has watched countless people undersell themselves. Apply for jobs even if you're uncertain you'll land one; the interview and test process will show you exactly where you stand.
— Matt BeardOne warm referral saves more time than a perfect application
Screening thousands of résumés is brutally costly, and even interviews and tests don't guarantee confidence in fit—so a referral from someone credible saves organizations enormous first-screen time. Even a weak signal like ‘we chatted at an event’signals seriousness and community engagement. But it's not required, the author stresses: many people get offers through cold applications, and most hiring processes include anonymous assessments to independently evaluate candidates.
— Matt BeardMoving to a hub city will teach you more than remote preparation
People who successfully moved into AI safety told the author that relocating to where things actually happen—San Francisco, London, Washington—was their most important decision. The embedded knowledge and professional culture in a local community can't be replaced by articles or podcasts. If moving isn't feasible, attend in-person conferences like Effective Altruism Global and treat conversations as research, not social obligations.
— Matt BeardPublic work gets recruiter attention much faster than private experience
In interviews with approximately 40 AI safety organizations, respondents stressed that public writing, visible projects, and shipped work matter more than accumulated quiet experience—recruiters can't see inside your head, but they notice details like ‘she published a response to our paper.’ Public work also accelerates feedback: releasing early for criticism saves months of wasted effort, and even harsh feedback reveals whether you're missing skills or just need clearer communication.
— Matt BeardYour side project can become your job if it solves real needs
The author describes a real case: Luca noticed AI safety lacked a talent networking platform, so he built one with login, opportunity matching, and recruiter support. Later, when 80,000 Hours was building a similar job board, someone suggested talking to Luca. After a few rounds of evaluation, they hired him full-time. The lesson: don't wait for a job offer to start doing the work you want. Even part-time side work gets you noticed faster.
— Matt BeardYour impostor syndrome probably stems from using the wrong comparison group
The author suggests three tests: Are you comparing yourself to people with 3× your experience or 5 years your senior? Is your reference group only high-profile leaders, not the broader ‘competent people at the frontier’? Have you genuinely tested your learning potential—invested hours, tried methods, gathered feedback? He also warns that burnout is real, and consistent daily practice beats sporadic bursts of effort.
— Matt BeardIn their own words · checked verbatim
I'd say you don't have to be certain, but you should have something to say.
Matt Beard2:43
It's fine if your career looks like this for a while. You should build a ladder of cheap tests to discover what you're good at.
Matt Beard4:06
The AI safety field has grown by about 3000% in the last eight years, driven by the urgency of these problems and a large talent development ecosystem.
Matt Beard5:49
you should treat your networking like a research project and you should try to remember that talent markets are shockingly inefficient.
Matt Beard7:26
Multiple people who successfully found work in AI safety have told me that moving to where the action is was their most important decision.
Matt Beard10:44
Getting feedback is scary, but essential.
Matt Beard12:20
Your goal isn't just to get a job, of course, it's to build the skills to solve the problems that you care about solving, and this is one way to do it if you have extra time.
Matt Beard15:35
30 minutes a day sustained over a few months will take you much farther than a couple frenzied weekends where you work on something and then give up.
Matt Beard15:35
Figures
| AI safety ecosystem growth (8 years) | approximately 3000% | 5:49 |
| AI safety organizations surveyed | approximately 40 | 12:20 |
Glossary
- BlueDot course
- Introductory online course series for AI safety and governance.
- EA Forum
- The effective altruism community's discussion forum, where AI safety researchers post.
- LessWrong
- Rationalist community forum where AI safety professionals actively discuss.
- deliberate practice
- Focused training at the edge of ability with active feedback loops.
- fellowship
- Short-term paid transitional position in a specific field, often a pathway to full-time work.
How to listen
Engineers, students, or career changers preparing to move into AI safety but unsure about the first step.