Fellowship
Applications are now open for our intro fellowship.
We offer two parallel introductory fellowships: AI Safety Fundamentals and AI Governance.
The fundamentals track introduces the technical challenge of making AI systems reliably follow human intentions, while the governance track examines the role of policy, institutions, and global coordination to reduce AI risks. Both cover forecasting how the technology develops.
Fellowships run weekly for 6 sessions, paper discussions over free, fancy dinner at Trajectory Labs, an off-campus AI safety hub.
Our fellows
I participated in a fellowship last fall, and I absolutely loved it! The fellowship gave me a friendly and passionate environment in which to explore recent research in AI alignment techniques during meals with other students. Since the fellowship, I've continued to develop my skills alongside these students, and have become much more informed and capable of working to improve AI safety.
BoyanFellow '25
Going in, I had some interest in AI safety but little idea how it shows up in real research or how someone technical like me could contribute. The curriculum and weekly discussions gave me a much clearer sense of the field, and I enjoyed the sushi.
DivyFellow '25
I came in curious and found a community of people who genuinely care about getting this right, a real grip on the technical landscape, and a clearer sense of where I want to contribute. The modern discussion space and free food are also awesome perks. These fellowships have given me a foundation for thinking about AI safety that I carry into everything I work on.
PeraFellow '25 and '26
Curriculum adapted from BlueDot Impact.
AI Safety Fundamentals
6 weeks
Topics include:
- Intro to deep learning (first session only)
- Forecasting
- Reinforcement learning from human feedback
- Scalable oversight
- Mechanistic interpretability
- Technical governance
- Contributing to technical AI safety
AI Governance
6 weeks
Topics include:
- Forecasting
- Overview of key actors
- Identifying levers for effective policy frameworks
- Governance at frontier labs
- Canada’s role in international cooperation
- Contributing to AI governance
FAQ
What if I don't have a technical background?
None of our fellowship streams require a technical background. Both will introduce you to the field of AI safety, with a focus on technical and governance aspects depending on the stream. We’ll focus mostly on the conceptual arguments for why we might expect misalignment and why it might be a hard problem. For the technical stream, we recommend watching 3B1B: But What is a Neural Network? to get some high-level intuition for how deep learning works. There won’t be any coding though.
What is the expected time commitment?
Around 3 hours per week. There will be around 1-1.5 hours of readings per week, and a 1.5 hour dinner meeting with your cohort to strengthen your understanding of each week’s material.
Where will the dinner meetings be held?
At Trajectory Labs, close to King station.
When does the reading group start?
The fall iteration will begin the week of September 20th.
What day/time do sessions run?
TAISI will be running multiple cohorts and you’ll be placed in the cohort that best matches your schedule. We can’t guarantee to accommodate everyone’s schedule, but we’ll try our best.
I have a question that wasn't answered here. Is there anyone I can reach out to?
For sure! Send an email to joseph@taisi.ca and we’ll get back asap.