Aran Nayebi

Message

I'm an Assistant Professor at Carnegie Mellon’s Machine Learning Department. I'm also a core faculty member in CMU’s Neuroscience Institute, and hold a courtesy appointment in the Robotics Institute.

My lab works at the intersection of neuroscience & AI to reverse-engineer animal intelligence and build the...

3mo

Aran Nayebi

From Barriers to Alignment to the First Formal Corrigibility Guarantees

This post summarizes my two related papers that will appear at AAAI 2026 in January: * Part I: Intrinsic Barriers and Practical Pathways for Human-AI Alignment: An Agreement-Based Complexity Analysis (selected for oral presentation) * Part II: Core Safety Values for Provably Corrigible Agents What these papers try to quantify...

Dec 8, 202561

This post summarizes my two related papers that will appear at AAAI 2026 in January:

Part I: Intrinsic Barriers and Practical Pathways for Human-AI Alignment: An Agreement-Based Complexity Analysis (selected for oral presentation)
Part II: Core Safety Values for Provably Corrigible Agents

What these papers try to quantify are two questions:

What factors make alignment hard in principle, and
Which safety targets remain tractable despite those limits.

The first paper gives formal lower bounds on the difficulty of AI alignment that apply even in best-case scenarios, yielding the first “No-Free-Lunch” theorems for alignment, namely: (1) aligning to “all human values” is intractable, and (2) reward hacking is inevitable in large state spaces & bounded agents.

The second paper shows... (read 3290 more words →)

AI ALIGNMENT FORUM
AF

AI ALIGNMENT FORUM
AF

Aran Nayebi

Aran Nayebi

Aran Nayebi

From Barriers to Alignment to the First Formal Corrigibility Guarantees

Aran Nayebi

Aran Nayebi

Aran Nayebi

From Barriers to Alignment to the First Formal Corrigibility Guarantees