The Future of Food: from AI to Urban Farms

This piece is part of the Ecofuturism Project, a series of articles looking at how new technologies can contribute to a more eco-friendly society.

By Sulaiman Ilyas-Jarrett

Food is important. In the developed world, we often take it for granted, yet the future of food is by no means certain. Current practices in agriculture and fishing are increasingly unsustainable. Food production already accounts for up to a third of our global carbon footprint, and poor resource management is eroding topsoil quality, depleting water reserves, and polluting ecosystems with harmful pesticides.

Something has to change. Past Future Society articles have looked at the need to shift from a meat heavy diet to a plant-based one, but here we will focus on creating better systems for food production. New methods and technology – such as Aquaponics, Agroforestry, and Artificial Intelligence – present the opportunity to radically reform the way we produce food, forming models much more in tune with our environment.

Hydroponics – the technology that’s got NASA excited

One technology attracting the interest of innovative food producers is Hydroponics: the practice of growing plants in nutrient-rich water instead of soil. While the method is not itself new, allegedly first used in the ancient Hanging Gardens of Babylon, it has been significantly refined in the last few decades, in part thanks to research at NASA, seeking to develop efficient closed food systems that don’t require soil.


Hydroponics systems have key environmental benefits. They use about 10% of the water that traditional agriculture uses, because water is recycled in a closed system and regularly topped up with nutrients. Freedom from soil means it is possible to use land much more efficiently as well, with plants often stacked vertically. This also allows food production to be brought closer to urban centres. From a restaurant in New York that grows all its produce on its rooftop, to the 150 ‘plant factories’ in Japan, and a vast new underground lettuce farm beneath the streets of London, this technology makes cities less reliant on rural imports. As well as lowering transport costs and CO2 emissions, this will make cities more resilient to extreme weather – which will become increasingly important as our climate changes. What might once have seemed like science fiction has fast become a viable, efficient, and profitable solution to food production.

Artificial Intelligence – can it work for the environment?

Asides from physically different systems, AI and automation will also play a pivotal role in creating better food systems. The most obvious way AI can help is by optimising the use of resources and reducing waste. The UN Food and Agriculture Organization estimates that over a fifth of agricultural production is wasted globally, and the overuse of fertilisers has led to nutrient run-off, disrupting aquatic eco-systems and causing ‘Dead Zones’ where they end up in the oceans. While some of this waste happens in the household, the majority occurs before it’s bought by a consumer. High tech farming (or ‘precision farming’ as it’s often known) has been around for a while now, with GPS systems that map terrain, driverless tractors, and increasingly drones that perform an array of tasks. Established companies like CNH Industrial have been driving the use of Big Data, but big firms have often been hesitant in their use of AI. It has been left to smaller start-ups, like Flux Farms or Observe Technologies, to focus on better software solutions.

It is notable that these start-ups work in hydroponics and aquaculture respectively – both closed systems which allow farmers more control over the production environment. This is important because the revolutionary new systems we need will combine multiple innovations. Rather than simply slapping AI onto an existing model, they will favour systems that have greater scope for improvement, such as hydroponics. AI may also help to manage systems that are difficult for humans to get right. A good example is Aquaponics, a mixture of hydroponics and aquaculture, where fish are added to the water cycle and their waste infuses it with nutrients. This has been difficult to scale because it adds more variables and therefore more complexity, but a well-designed AI system could account for these and keep the system in balance, especially given it occurs in a controlled environment.

Artificial Intelligence is not, however, intrinsically positive. Simple machine learning can optimise existing systems, but unless we use it in conjunction with new models it may not produce the desired results. As we move towards more sophisticated AI models, we must be clear what it is we want from these algorithms. Their goals should be to optimise environmental objectives, not necessarily yields or profit. Doing so means accurately measuring and quantifying the impact of farming on the environment, which is no easy task. The next few years will be crucial in defining the role of AI in food production. Policy-makers may not want to leave this entirely to the private sector. Government guidelines, and potentially regulation, will help promote this vital long-term thinking.

Permaculture and Agroforestry – making the most of ‘natural systems’

Finally, we must not neglect the role of natural systems. Permaculture and Agroforestry are two methods that aim to work with nature, rather than against it, to grow food more sustainably and ultimately produce higher yields. The key principle behind both is to create or foster self-sustaining ecosystems, which limits the need for labour and other inputs. While not exactly high tech, both are examples of innovative system design. Farmers must take into account the ecology, topography, and sometimes nano-biology of their site, in order to design a system that fits the specific needs of the land.

Both techniques are gradually gaining attention, not only for their ability to regenerate over-exploited land or prevent deforestation, but also because they often generate higher yields in the long run. However, many farmers, interested primarily in short term profits, have little incentive to switch to these methods. Though there is a grassroots permaculture movement in Europe and the US, these methods are mostly deployed in places where land is already damaged, or deforestation is particularly acute: Brazil being an example where both issues are at play. This is another area where governments could show leadership. Subsidies are already a staple of agricultural policy in the US, EU, and beyond – tying those subsidies more directly to environmental impact would be a good place to start. Favourable loans to assist with the high set-up costs would also promote these innovative methods.

To Conclude

Achieving sustainable food production requires holistic and long term thinking. It is not enough to tinker around the edges of existing practices. The systems described in this article embody that, and provide some alternative models. While much of this innovative technology is being developed in the private sector, governments must take an active role in promoting better food systems. Only bold, forward-thinking policy can stir systemic change.


Sulaiman Ilyas-Jarrett is an incoming Isaac Newton Scholar in Environmental Policy at the University of Cambridge. He is fascinated by the role of innovation in promoting social change. He has worked on several UN projects looking at technology and governance, as well as with some of the UK’s leading social enterprises and start-ups; including Bulb, Up Learn, and Observe Technologies.