Why does 80,000 Hours focus on AI risk, and what can I do about it?

Updated

80,000 Hours focuses on AI risk because its research concludes that helping AI go well is currently one of the highest-impact ways to spend a career. The site states that many experts believe there is a 1-in-10 chance (or worse) that AI could cause human extinction, and that far more people — with both technical and non-technical backgrounds — are urgently needed to work on the problem. If you accept that premise, the practical question becomes where you fit: understanding the risks, finding a role, or building skills first.

The reasoning behind the focus

80,000 Hours frames career impact around the problem you work on, which it calls probably the largest factor in how much difference you can make. It selects problems using three criteria:

  • Scale — how many people (or how much of humanity's future) is affected
  • Tractability — whether additional effort can actually make progress
  • Neglectedness — whether the issue is unduly ignored by the rest of society

AI risk scores highly on all three in their assessment, which is why it sits at the top of their mission. The site also notes that some problems may threaten humanity's future, making them especially high priority.

What 80,000 Hours says the risks are

The site presents an uncertain ranking of global problems where additional people can have the most impact. The AI-related entries are:

Problem The concern as described
Loss of control We may develop very powerful AI systems before adequate control mechanisms exist. If such systems accumulate power while misaligned with human values, the result could be catastrophic.
Extreme power concentration AI could turbocharge efforts to seize and hold power. Surveillance, persuasion, and control technologies could let small groups or authoritarian leaders entrench their rule, potentially with little chance of recovery.

Other problems in their ranking include engineered pandemics, great power conflict, and factory farming — so AI is presented as a leading priority, not the only one.

What you can actually do

80,000 Hours describes three things its research, podcast, and advisors are designed to help you do:

  1. Understand AI and the key risks — build enough familiarity with the arguments to judge them yourself rather than taking the 1-in-10 estimate on faith.
  2. Figure out where you fit — match your existing skills and background to a role, whether technical or non-technical.
  3. Skill up, or find your role right away — either prepare for a future position or apply for something available now.

The site says it has researched AI for the last decade and given personal career advice to thousands of people, and that you can access all of its advice for free. It also invites visitors to sign up to hear from it as a starting point.

How to decide whether this applies to you

This framing is most useful if you are choosing what problem to work on and are open to AI safety or governance as a career focus. It is less directly relevant if you have already committed to a different cause area, or if you disagree with the underlying probability estimates — the site itself labels its problem ranking as uncertain, so treat it as an argument to evaluate rather than a settled fact.

A reasonable first step is to read the "loss of control" and "extreme power concentration" pages, then ask whether your current skills map onto any of the roles the site describes. If they don't yet, the skill-up path is the intended entry point.