As the field of Artificial General Intelligence (AGI) safety grows, there's increasing urgency to understand why talented individuals face difficulties entering the field, becoming productive contributors, forming collaborations, or sustaining long-term engagement. Without identifying these bottlenecks, community-building efforts risk being misdirected, potentially slowing progress in this time-sensitive domain.
One approach to uncovering these bottlenecks could involve conducting structured interviews with three key groups:
The interviews would focus on challenges faced, gaps in current support systems, and missing forms of community support. The findings could then be synthesized into a report highlighting the most severe bottlenecks along with concrete suggestions for addressing them.
A phased approach might work best:
To validate the approach, one could compare self-reported versus observed bottlenecks in a small sample, examine whether challenges are consistent across career stages, and assess the feasibility of proposed solutions with program organizers.
This approach differs from existing efforts in several ways:
The findings could help program organizers improve their offerings, assist new researchers in navigating the field more effectively, and guide funding organizations toward the most impactful community-building interventions.
While challenges like sampling bias and social desirability effects would need to be addressed through careful study design, this approach could provide much-needed clarity about how to most effectively grow the AGI safety field.
Hours To Execute (basic)
Hours to Execute (full)
Estd No of Collaborators
Financial Potential
Impact Breadth
Impact Depth
Impact Positivity
Impact Duration
Uniqueness
Implementability
Plausibility
Replicability
Market Timing
Project Type
Research