RSS

Timing of ex­is­ten­tial risk mitigation

TagLast edit: 17 Jan 2022 13:34 UTC by Pablo

Some risks, like climate change, are likely to have only moderate effects in the near term but potentially large ones in the long term. Others, like asteroid risk, are roughly evenly distributed over time. Even within risk areas, the mechanism that creates the risk can be different at different time periods. For example, if climate change does create large impacts in the next decade, it is likely to be because of unforseen feedback effects, rather than because of accumulation of warming from industrial emissions.

In all of these cases, the work that one should do to mitigate risk will differ depending on which time-frames one is considering. A balanced portfolio of risk-mitigation strategies therefore needs to allocate resources against these different timelines.

Two separate considerations favor a focus on near-term risk. First, we are short-sighted about the nature of future risks and what can be done about them, so our efforts are better targeted when acting on near-term risks. Second, we are the only ones who can act now, whereas future risks can be addressed by others later.

On the other hand, risks that emerge in the future might be larger. Moreover, working now on building capacity to address risks in the future might help scale up the total amount of effort going to risk reduction.

The optimal strategy for a risk community is likely to be a mixture of work targeting different time-horizons. There are some basic quantitative models suggesting how to address this balance.[1]

Further reading

Cotton-Barratt, Owen (2015) Allocating risk mitigation across time, technical report #2015-2, Future of Humanity Institute, University of Oxford.

Ord, Toby (2014) The timing of labour aimed at reducing existential risk, Future of Humanity Institute, July 3.

  1. ^

    Cotton-Barratt, Owen (2015) Allocating risk mitigation across time, technical report #2015-2, Future of Humanity Institute, University of Oxford.

No entries.