If AI Dethrones Us, What Will It Do with the Natural World?
The usual topic of discussion, our love lives, only holds up about as long as Joel Usher’s ability to suppress his passion for Artificial Intelligence. For him, the two topics are merged. Once, as we sat surrounded by history and tradition in an Oxford pub, Joel explained that AI could soon usurp humans as the dominant player on Earth, highlighting that some of the world’s most influential people (Stephen Hawking, Elon Musk) have signalled warnings.
While Joel is far more convinced of AI’s upside, such as the release of humanity from the constraints of labour, he summarised for me how it could come to pass that AI supersedes human intelligence, identifies us as a threat, risk, or simply inferior, and subsequently neutralises us. I questioned Joel on what might happen next. We had a brief think about this before returning to the discussion of our love lives…with women.
Since, I’ve been wondering, “Alright, we’re some time on, and AI has done away with humans. So it’s just AI and the natural world. Now what.”
Backtracking a bit, one stream of logic presented for why AI might dispense with us in the first place is that we have proven to be a destructive force upon the Earth. We live in an era that scientists call the ‘Anthropocene’, a term for which these same scientists can’t agree upon a definition, but in any event we’ll take it to mean that humans have fundamentally altered the way the planet functions. For example, whereas climate change has been driven by natural circumstances in the past, we now drive it, placing life as we know it on earth in jeopardy.
So, from the natural world’s perspective, it could be AI to the rescue, removing the bad guys, us, to set things right. Should this be the case, let’s look at a couple of approaches that AI might take:
- Hands-Off: A popular movement in recent years calls for the rewilding of the natural world. The approach suggests that nature should be allowed to find its own balance according to its own processes. Within this movement, we humans are also encouraged to rewild ourselves, but please recall that at this point the machines have killed everyone. Now, if AI were to opt for rewilding, it still has some decisions to make. It might seek to return nature to a baseline, such as to a time of pre human-driven extinctions, reintroducing extinct megafauna in the process. AI might also choose to go hands-off hands-off, allowing nature to continue from the point of AI takeover, and choosing to accept lingering human effects as consequences of our existence.
- Hands-On: Moving to the other end of the spectrum, AI might choose to micromanage all things natural. It could perhaps do away with Protected Areas, given that they were created by humans to protect nature from humans, who of course are all now gone. Or might it determine that nature has its place only where deemed appropriate, as we’ve done, allocating space while seeking to ensure that nature is contained therein? Interestingly, AI would also be left to determine whether or not nature requires protection from its own forces. Might AI intervene when an invasive species comes knocking to disrupt a steady ecosystem? Would it geo-engineer the climate to maintain a stable state? Would it call forests to order, demanding perfect rows and columns and good behaviour?
Of course, these discussions fall within the realm of present human debate, but it’s altogether likely that the decision criteria change. Experts believe that AI might undertake a mass optimisation process, packing the planet full of computing power. In this instance, the natural world becomes a space constraint. Moreover, while at present we haven’t found replacements for the ecosystem services that keep us alive (ie. food, water, clean air, Sir David Attenborough), super-intelligent AI could very well figure out how to replicate these services via artificial means. In fact, AI mightn’t need these services at all.
Where does this leave us then?
The topic calls into question what it is that we can learn from an AI approach to nature, whether that approach would be to let nature do its thing, manage its conservation more closely, or replace it with something artificial. I largely support the first option, with, obviously, us still in the picture. But we now touch nearly everything on earth in some way, and over time we leave less space and scope for the natural world to conduct itself free of our influence. An artificial intelligence takeover might, oddly, be the only way for truly natural occurrences to re-establish themselves.
One of the world’s leading thinkers on AI and Oxford professor, Nick Bostrom, frames our challenge as follows:
I believe that the answer here is to figure out how to create superintelligent A.I. such that even if — when — it escapes, it is still safe because it is fundamentally on our side because it shares our values. I see no way around this difficult problem (TED2015).
…Difficult problem, indeed. Considering that we have leaders of nations and others who seek leadership denying the science around climate change, the installation of shared values appears to be a bit of a stretch. But we are unique in our values systems. Our ability to care for each other and our natural surroundings is nonetheless a powerful, intrinsic force, and one that AI mightn’t ever grasp. Yet, given that those at the forefront of AI warn that super-intelligence is not as far off as we might think (~2040), we had better come to terms collectively on what the natural world means to and for us. Otherwise, an AI-owned future could prove better suited to co-existence with the natural world than one owned by us, or it could prove that co-existence of any kind simply isn’t in the plans.
Certainly, in this case and all others, we are responsible for the future of life forms other than our own. On many fronts, we must act to keep it that way, and more importantly, we must act to deserve to keep it that way.
— — — — — — — — — — — — — — — — — — — — — — — — — — — — — — —
Top 5 Links
Joel Usher’s post on AI and the future of work.
The Future of Life Institute’s open letter on AI.
The Oxford Martin School’s Future of Humanity Institute.