In Australia, autonomous killer robots are invading the Great Barrier Reef. Their target? The crown-of-thorns starfish, a malevolent pincushion with a voracious appetite for corals. To protect ailing reefs, divers have often injected the starfish manually with bile or vinegar. But this manual activity can now be carried out by a robot called COTSbot. Developed by a team of Australian scientists, this intelligent underwater robot has learned to identify the starfish among the coral, and can execute them by lethal injection. This technological escamotage might not be able to save the reef from erosion, but is useful to ponder on the approach to this issue. The work of conservationists is typically directed at reducing human influence, which is deemed harmful in most cases: from breeding the species we’ve killed to removing the pollutants we’ve added, and so on. All of these measures always involve human action. With the COTSBots, things are different: developed out of human intelligence, they’re designed to ultimately operate without us. They are the symbol of an increasingly frequent movement to save ecosystems removing human influence entirely.
In an intriguing intellectual experiment, landscape architect Bradley Cantrell, historian Laura Martin and ecologist Erle Ellis have taken this theory to its extreme consequences, ending up with a “wildness creator”, a hypothetical artificial intelligence designed by the human mind yet then left free to develop its own strategies for protecting nature autonomously. Although this solution is technologically, politically or financially impossible, it makes us ponder on what lies behind the human desire to save natural places. In trying to maintain natural places, we often end up over-curating them: the Australian COTSBots are a perfect example. In fact, according to ecologist Ellis, the crown-of-thorns starfish of the Australian coral reef isn’t even an invasive species. When we use artificial intelligence to preserve our planet, the inputs and motivations are always human-centric. But what would happen if this were not the case? Artificial intelligence is making huge progress, machines are capable of developing their own behaviour, going beyond that for which they have been originally programmed. An illustrative example: when Google’s AlphaGo system recently beat the Chinese champion Ke Jie in the Go game with an unconventional move that no human had ever made, our certainty as to the supremacy of human intelligence wavered for a moment. “After humanity spent thousands of years improving our tactics – said reigning human champion Ke Jie – computers tell us that humans are completely wrong”.
Several geoengineering projects are moving in this direction: from drones that can plant trees to artificial pollinators, up to swarms of oceanic vehicles for cleaning up oil spills. The recent trend is to cordon off portions of the Earth to protect it from human influence: if we use technology to reach this goal, we might end up with a world in which several ecosystems will exist without us. The idea of removing ourselves from nature is unachievable; rather, we should think about finding a way to co-exist without stamping only humanity onto wilderness. How technology will be used in this scenario will prove to be strategic: will it be able to favour a non human-centric co-existence? Will it contribute to safeguarding and protecting what is not human and is in line with our rules? The real issue, as always, is in our hands.