Are policies that address x-risks and current harms from AI opposed?

8 min read

Suggest changes in Google Docs

There is a perceived tension between emphasizing the current harms from AIs (such as discrimination and misinformation)¹ and emphasizing the existential risks from this technology. This tension is sometimes used to argue for focusing exclusively on one of these types of risk. This tension is often exaggerated and we argue here that there are good reasons for people working on these issues to work together rather than separately.

Several factors contribute to the existence of this perceived tension:

Limited attention and resources
Different cultures
Differing priors on the likelihood of extinction from AI
Uncertain effects of regulation

Limited attention and resources

When policy makers and lay people are exposed to “risks from AI”, some have argued that these immediate and future risks might be bundled together, and that existential risks might take up all the air in the room, leading to ignoring other issues from AI.

Is this argument correct? In the general case, no. If it were true, then by analogy we could argue that any two causes would compete against each other for resources. But no one would claim that working on saving the environment sucks up oxygen for trans rights activism. So there must be something special about the case of current AI harms and AI x-risks which allows this argument to work.

future harms from AI, we observe that this tension is more present than between say environmental activists and trans rights activists.

[To our knowledge, there are two reasons this might be true. Firstly, that the two causes sound similar enough that people hearing them for the first time would confuse them which would cause people to confuse solutions for one for the other. And secondly that if AI x-risks, if real, would completely overwhelm existing harms in people’s minds. ]

[We needs to actually address these two points]

But in practice, talk of existential risk has not been detrimental to the discussion on current harms. Others have found that multiple metrics indicate that the prominence and funding of people and organizations that concentrate on current harms has not been affected negatively by the sharp increase in interest for existential risks in the spring of 2023. Others still point out that this vision of attention as a zero-sum game is misleading.

https://www.pnas.org/doi/10.1073/pnas.2419055122

https://link.springer.com/article/10.1007/s43681-023-00336-y

Different cultures

Historically, the concern about discrimination has been brought up by academics close to the humanities and concerned by the plight of people being marginalized by this discrimination.² By contrast, concern for existential risks was initially promoted by people associated with transhumanism³ concerned for the future of humanity as a whole.⁴ ⁵ While in theory these two communities could work together and sometimes do, tribalism has contributed to keeping some distance between them.

Differing priors on the likelihood of extinction from AI

Although it is not always named, people who oppose legislation to reduce x-risks from AI tend to believe AGI won’t happen soon, and tend to believe it won’t be dangerous.

As an analogy, consider climate change. If you think climate change is a big problem, and imminent, then you’d really want to work on ameliorating climate change’s harms. If you disagree with those premises, then you won’t want to make sacrifices such as taxing people’s environmental impacts. Given their assumptions, both stances make sense. And in fact, many people do have one assumption or the other. You might’ve even heard someone mention it on the news. But that doesn’t mean there’s no room for collaboration! Cheap solar power is good no matter the urgency of climate change.

In a similar fashion, people who are skeptical of x-risk from AGI or believe that AGI is either impossible or very far off will understandably want to avoid paying opportunity costs to avoid a harm that they do not believe will happen anytime soon or at all. It is likely that if such skeptics became more convinced of the seriousness and imminence of such risks, they might change their priorities.

Uncertain effects of regulation

There is some disagreement within people who are concerned about x-risks as to whether stringent regulation such as a pause on AI or restricting access to advanced chips is beneficial to safety or not. If we are not reasonably certain that enacting such policies will result in a safer world, it is less tempting to use political capital to push for such policies at the expense of others where we are more confident in the effect.

Whilst it is true there’s great disagreement over how or why AI could pose an x-risk, there are still policies which would be work under most world-views. For instance, a global ban on developing lethal autonomous weapons further. Or a liability on AI labs which would internalize the negative externalities of their actions.

Reasons to work together

On the flip side, there are reasons for these communities to work together.