# Call for Talent: Learning-Theoretic AI Alignment Mentorship at PIBBSS

> Coverage of lessw-blog

**Published:** January 14, 2026
**Author:** PSEEDR Editorial
**Category:** risk
**Content tier:** free
**Accessible for free:** true



**Word count:** 385


**Tags:** AI Alignment, PIBBSS, Recruitment, Theoretical Computer Science, Learning Theory, AI Safety

**Canonical URL:** https://pseedr.com/risk/call-for-talent-learning-theoretic-ai-alignment-mentorship-at-pibbss

---

A new recruitment drive seeks mathematicians and theoretical computer scientists to address the technical challenges of AI alignment through the PIBBSS summer fellowship.

In a recent post, lessw-blog announces a specific mentorship track within the upcoming PIBBSS summer fellowship, aimed at researchers with strong mathematical backgrounds interested in the Learning-Theoretic AI Alignment (LTA) agenda.

**The Context**

As the capabilities of artificial intelligence systems accelerate, the field of AI safety has moved beyond philosophical debate into rigorous technical research. A core challenge remains the "alignment problem"—ensuring that superintelligent systems act in accordance with human intent. Within this broader field, various technical agendas have emerged, each proposing different mathematical or theoretical frameworks to guarantee safety. Finding talent capable of engaging with these high-level theoretical problems is a significant bottleneck. The LTA agenda represents a specific, mathematically grounded approach to these challenges, requiring researchers who can navigate complex learning theory and computer science fundamentals.

**The Gist**

The post serves as a direct call for applications for the 2026 PIBBSS (Principles of Intelligent Behavior in Biological and Social Systems) summer fellowship. Specifically, it highlights mentorship under Vanessa, a researcher focusing on the LTA agenda. The author frames this specific research direction as a critical, high-stakes pathway to solving the technical alignment problem before the potential emergence of artificial superintelligence.

The recruitment is highly targeted. Unlike general policy or governance roles, this track seeks candidates with deep expertise in mathematics and theoretical computer science—typically PhD students or postdocs. The post emphasizes that the LTA agenda has "shovel-ready" research directions, suggesting that the theoretical groundwork is established and skilled researchers are now needed to execute proofs and expand models. Furthermore, the post notes that exceptional fellows may be funneled into longer-term, paid research fellowships at CORAL, indicating a structured talent pipeline for this specific technical agenda. The author stresses that applicants should be seriously interested in dedicating their careers to AI alignment, positioning this opportunity as a gateway to full-time research in the field.

For those with the requisite mathematical background, this post offers a concrete entry point into high-impact safety research.

[Read the full post](https://www.lesswrong.com/posts/NGG9NZFpwiRBFyS4b/apply-to-vanessa-s-mentorship-in-pibbss)

### Key Takeaways

*   Applications are open for the PIBBSS summer fellowship 2026, featuring a specific track for Learning-Theoretic AI Alignment (LTA).
*   The mentorship targets PhD students and postdocs in mathematics and computer science.
*   The LTA agenda is presented as having 'shovel-ready' problems requiring immediate technical attention.
*   Successful fellows may be offered paid research positions at CORAL.
*   The post frames this work as critical for mitigating existential risks associated with unaligned superintelligence.

[Read the original post at lessw-blog](https://www.lesswrong.com/posts/NGG9NZFpwiRBFyS4b/apply-to-vanessa-s-mentorship-in-pibbss)

---

## Sources

- https://www.lesswrong.com/posts/NGG9NZFpwiRBFyS4b/apply-to-vanessa-s-mentorship-in-pibbss