Why does AGI matter?

What benefits could AGI bring?

What risks does AGI pose?

Terrorism

Authoritarianism

Loss of Control

Summary

You can make a difference

How our courses work

Getting started

What's already possible

The pace of progress

From tools to agents

What is AGI?

Vending-Bench: measuring progress towards AGI

Building AGI and beyond

The huge bet on AGI

You've made it through three units on what AI can do, where it's going, and what's at stake. You might be feeling the obvious next question: what now?

The honest answer starts with a number. We estimate there are only a few thousand people working full-time on frontier AI safety. Compare this to the millions of people, quite rightly, working other global challenges such as averting climate change.

A few thousand people is woefully insufficient to address the scale of this challenge.

But it's also an opportunity for you. The problem has many parts; technical, political, scientific, organisational, and the field is small enough that you can become a world expert in a specific corner of it in a matter of months. You don't need a computer science PhD, you just need to start.

At BlueDot, we fundamentally believe that the decisions humans make will determine the trajectory of this technology, and we want to support and empower you to shape that for the better.

We've helped thousands of people understand where AI is going and what they can do about it. Hundreds now work directly on these problems: in AI company safety teams, think tanks, governments, and new organisations they founded themselves. We run [courses](https://bluedot.org/courses), [events](https://luma.com/bluedotevents) around the world, and give [grants](https://bluedot.org/programs/rapid-grants) to get people building.

The rest of this unit walks through how to find the path that fits you.


Our main courses take the ideas introduced over the past few units and go much deeper.
- The [AGI Strategy Course](https://bluedot.org/courses/agi-strategy) covers the technical and economic drivers of AI progress, modelling threat pathways and the ways we can defend against them. We recommend you start here if you’re newer to AI Safety.
- The [Biosecurity](https://bluedot.org/courses/biosecurity), [Technical AI Safety](https://bluedot.org/courses/technical-ai-safety) and [Frontier AI Governance](https://bluedot.org/courses/ai-governance) courses dig deeper into specific sub-fields and the concrete ways you can contribute.

<Embed url="/images/content/foai/course-paths.webp"/>

The format is more intensive than this course was; 25-30 hours of reading and small-group discussions, led by a facilitator who is working directly in the field. You work through a structured curriculum of curated resources before debating your ideas off 5-8 other people who are as committed to working on these problems as you are.

<Embed url="/images/content/foai/cohort-call.webp"/>

**We're selective about who joins each cohort to keep the discussion quality high.** We accept applicants from a wide range of backgrounds, but the strongest applications make clear what experience you're bringing, where you want to apply it to make AI go well, and how the course will help you get there. Your plan doesn't need to be perfect (that’s what the course is for!), but you do need to have thought about it.

For top participants, the support continues after the course ends. We run [1-1 advising calls](https://bluedot.org/programs/advising) to think through what to do next, and we provide [grants](https://bluedot.org/programs/rapid-grants) to fund promising projects that come out of the course.

<Callout title="Not ready to apply yet, or didn't get in?">
There are many lower-touch ways to engage with AI safety, that will also strengthen your application to future rounds of the courses
- **Self-study the course materials.** All of our curricula are openly available on the course pages. Our top recommendation to improve your course application is to read through some of the materials, write up your thoughts and integrate this into your next application.
- **Find a local AI safety community.** There are hundreds of AI safety communities around the world ([map here](https://aisafety.com/communities)) if you'd like to meet others working on these questions.
</Callout>



You've made it through the course! Before you close the tab, take 5 minutes to note down what you'd tell a friend over coffee if they asked what you took from it.

The shape we suggest:
1. **What stood out to you.** Pick one or two ideas from the course that feel the most juicy or important to you.
2. **What you bring to the table.** Step back from AI for a moment. What have you spent time getting good at, in your work, your study, or in your side projects? Which bits of that experience could matter for making AI go well?
3. **Where you could go from here.** Given the first two answers, what's a direction you'd want to explore?

Your answers to those three questions are probably a good starting point for an [AGI Strategy Course](https://bluedot.org/courses/agi-strategy) application!


Future of AI

Getting started

Exercises