# How Do AI Timelines Affect Existential Risk?

**Authors:** Stephen McAleese  
**arXiv:** [2209.05459](https://arxiv.org/abs/2209.05459)  
**Date:** 2022-08-30  
**Summary page:** https://www.summarizepaper.com/en/arxiv-id/2209.05459v1/

## Key points

- Potential creation of superhuman artificial general intelligence (AGI) in this century
- Implications for existential risk
- Delaying AGI development could decrease total existential risk by allowing more time to work on AI alignment problem
- Acknowledges that delaying AGI could increase other existential risks, such as those posed by advanced technologies like synthetic biology and molecular nanotechnology
- Focus on AI timelines and their impact on existential risk
- How quickly AI alignment work could reduce existential risk from AGI
- Consideration of initial level of AI existential risk
- Examination of relationship between AI risk and other existential risks (natural state risks, anthropogenic state risks, anthropogenic step risks)
- Introduction of factors influencing existential risk: war, cognitive enhancement, hardware overhang
- Discussion on potential effects of these factors on AI risk
- Conclusion: Delaying creation of AGI may decrease total existential risk if AI existential risk is high relative to other risks
- Emphasis on taking robustly positive actions to reduce overall existential risk:
- Research related to analysis of global catastrophic risks
- Developing strategies for effective governance and safety measures for artificial intelligence
- Promoting differential technological development to encourage responsible use while minimizing harms or dangers.

## Summary

The existing summary discusses the potential creation of superhuman artificial general intelligence (AGI) in this century and its implications for existential risk. It suggests that delaying the development of AGI could decrease total existential risk by allowing more time to work on the AI alignment problem. However, it also acknowledges that delaying AGI could increase other existential risks, such as those posed by advanced technologies like synthetic biology and molecular nanotechnology. The new context provided focuses on AI timelines and their impact on existential risk. It explores how quickly AI alignment work could reduce existential risk from AGI and considers the initial level of AI existential risk. The relationship between AI risk and other existential risks is also examined, including natural state risks, anthropogenic state risks, and anthropogenic step risks. Additionally, the context introduces other factors that can influence existential risk. These include war, cognitive enhancement, and a hardware overhang. The potential effects of these factors on AI risk are discussed in detail. In conclusion, the summary highlights that delaying the creation of AGI may decrease total existential risk if AI existential risk is high relative to other risks. It emphasizes the importance of taking robustly positive actions to reduce overall existential risk such as working on research related to analysis of global catastrophic risks; developing strategies for effective governance and safety measures for artificial intelligence; and promoting differential technological development which encourages responsible use of technology while minimizing associated harms or dangers. Overall, this refined detailed longer summary provides a comprehensive overview of the relationship between AI timelines and existential risk while considering various factors that can influence these risks.