Open Questions for the Near Future

The frontier raises unresolved questions about capability, safety, and governance. This sub-article expands Section 8 of Recent Topics.

Capability Questions

• Will “generalist” agents replace narrow tools?
• Can we reach verifiable AGI, and how would we know?

Safety Questions

• Can alignment keep pace with capability?
• How do we verify safety of general-purpose systems?

Governance Questions

• Who governs compute, data, and frontier models?
• How do societies adapt to ubiquitous synthetic content?

Social Questions

• How do humans and AIs share decision-making?
• What does flourishing look like in a world with powerful artificial minds?

Why It Matters

These questions are unresolved and will shape the next decade of AI.

Sources

• Bostrom, N. (2014). Superintelligence.
• Russell, S. (2019). Human Compatible.

Back to Recent Topics.