Time to Put Artificial Intelligence to Work for the Earth – AI for Good Summit
May 10, 2018
In 2018 everyone is starting to see the business value of AI. It is being added to more and more things every year, and the AI is getting smarter and smarter - accelerating human productivity and discovery. But as AI becomes more powerful and broader in its use and impact across industry, the unresolved issue of AI safety is paramount. The challenge, however, goes beyond guiding “human friendly AI” to ensuring “Earth friendly AI”.
Over the last 70 years human pressures on our planet have grown exponentially, placing our climate, water, air, biodiversity, forests and oceans under mounting strain. Scientists have identified nine critical ‘Planetary Boundaries’ - limits to things like ocean acidification, freshwater use, air pollution, biodiversity loss and climate change - which we must stay within if we are to continue to survive and thrive. Apparently we have already exceeded four out of the nine.
Against this backdrop, how can we harness the intelligence and productivity gains of the AI revolution to create the sustainability revolution we so urgently need?
In our recent research over 80 applications of AI were identified that tackled climate change, biodiversity and conservation, ocean health, water security, clean air and disaster risk. Our report also highlighted a timeline for some of the identified “game changer” AI applications for Earth challenges, from distributed energy grids to smart agriculture (see Figure). In many, AI comes together with other emerging technologies, including the Internet of Things, Blockchain, drones, and AR/VR.
More broadly as AI innovations that enhance computational power gather pace, such as deep learning chips, distributed computing and quantum computing, these developments will further scale and transform the opportunity to harness AI for the Earth. While classical computers cannot compute things the way nature does, advances in quantum computing will increasingly open doors for new scientific discoveries. They can be applied to quantum problems as they exist in nature and identify ways in which the Earth system really works, potentially leading to the discovery of new advanced materials for clean energy systems, or breakthroughs in climate and weather prediction.
Harnessing the full opportunity of AI for the Earth will require decisive action and collaboration. As the pressing fields of AI safety, ethics, impact, and governance progress over coming years, sustainability considerations must be a core component. Industry groups including the Partnership on AI will have an important role to play, alongside AI research labs, governments and international bodies like the United Nations.
Collaboration between AI experts, domain specialists in industry and academia, and authorities as well as philosophers and non-profits will also be key to helping us apply AI successfully to Earth challenges. As an example, the “4IR for the Earth” initiative– a collaboration between the World Economic Forum, PwC, and Stanford University – provides a platform to accelerate partnerships and projects between technologists, scientists, industry and governments that put AI to work for the planet.
Also, the second edition AI for Good summit, a partnership between PwC and ITU, brings together AI experts from academia and business with 25 UN organisations working together to design practical applications of AI to help achieve the SDGs. For example, projects that use satellite imagery together with machine learning to help meet Sustainable Development Goals 2 and 15 (Sustainable Agriculture, and Life on Land).
The good news is that we can now use AI to tackle some of our Earth’s most immediate and pressing challenges. The AI opportunity must be seen - by all stakeholders - as not only a business or economic imperative, but an Earth imperative.
Summary from a blog originally published in ITU magazine AI for Good Edition