Apply for the 2026 Affine Superintelligence Alignment Seminar | AI Safety Research with UC Berkeley
The Affine Superintelligence Alignment Seminar invites global researchers to address critical AI alignment challenges. Drawing on recent academic insights and open-source advancements, the seminar bridges theory and practical implementation in AI safety.
Apply for the 2026 Affine Superintelligence Alignment Seminar | AI Safety Research with UC Berkeley
summarize3-Point Summary
- 1The Affine Superintelligence Alignment Seminar invites global researchers to address critical AI alignment challenges. Drawing on recent academic insights and open-source advancements, the seminar bridges theory and practical implementation in AI safety.
- 2Hosted by nonprofit minily.org, this research-focused seminar builds on UC Berkeley’s Neyman Seminar and integrates open-source tools from the AFFiNE platform to ensure human values remain central in advanced AI systems.
- 3This is a selective research fellowship for professionals with demonstrated expertise in: Machine learning ethics and value alignment Formal verification of AI systems Human-computer interaction and AI behavior modeling AGI safety research or policy analysis Selected participants receive full funding, access to proprietary alignment datasets, and mentorship from leading AI safety researchers, including contributors to UC Berkeley’s Center for Human-Compatible AI and the AFFiNE open-source project.
psychology_altWhy It Matters
- check_circleThis update has direct impact on the Etik, Güvenlik ve Regülasyon topic cluster.
- check_circleThis topic remains relevant for short-term AI monitoring.
- check_circleEstimated reading time is 3 minutes for a quick decision-ready brief.
Apply for the 2026 Affine Superintelligence Alignment Seminar
The Affine Superintelligence Alignment Seminar has opened applications for its 2026 cohort, inviting PhD candidates, ethicists, and AI engineers to advance AI alignment and superintelligence safety. Hosted by nonprofit minily.org, this research-focused seminar builds on UC Berkeley’s Neyman Seminar and integrates open-source tools from the AFFiNE platform to ensure human values remain central in advanced AI systems.
Who Should Apply?
This is a selective research fellowship for professionals with demonstrated expertise in:
- Machine learning ethics and value alignment
- Formal verification of AI systems
- Human-computer interaction and AI behavior modeling
- AGI safety research or policy analysis
Selected participants receive full funding, access to proprietary alignment datasets, and mentorship from leading AI safety researchers, including contributors to UC Berkeley’s Center for Human-Compatible AI and the AFFiNE open-source project.
Research Partners: UC Berkeley & AFFiNE
The seminar’s framework draws directly from Aaron Roth’s March 2026 Neyman Seminar, "Agreement and Alignment for Human-AI Collaboration," which introduced formal models for consensus-building between humans and AI. Participants will analyze these models through adversarial simulations and case studies rooted in peer-reviewed papers.
Meanwhile, the AFFiNE v2026.3.13-canary.913 release demonstrates how alignment principles translate into practical software — from improved resource loading to verifiable rendering logic. These technical refinements aren’t just performance boosts; they’re safeguards against emergent misalignments in user-AI interactions.
Application Timeline & Deadlines
Applications open immediately and close on June 30, 2026. Unlike corporate AI summits, the Affine Seminar avoids promotional agendas. Its curriculum is built on open-source code audits, reproducible experiments, and transparent peer review — ensuring alignment research remains independent, rigorous, and accessible.
Why "Affine"?
The name "Affine" references mathematical transformations that preserve structure without distortion — a metaphor for maintaining human intent as AI systems grow more autonomous. In an era of rapid AI advancement, this seminar offers a rare, ethics-first pathway to shape superintelligence safety with integrity.
As AI capabilities accelerate, structured, transparent alignment efforts are no longer optional. Join the 2026 cohort and help define the future of human-AI collaboration.
