Skip to content

Climate & Energy

Beyond Renewables and Carbon Capture: How Governments Can Harness AI-Era Technology to Reimagine Climate Action


Commentary7th December 2023

The solution to delivering meaningful climate action lies with technology. As global leaders gather in Dubai for COP28, significant attention is rightly being given to core technologies such as renewable-energy generation, hydrogen production and carbon capture. But not enough attention is being paid to the potential impact the next technological revolution could have on enabling, improving and accelerating climate efforts, powered by artificial intelligence (AI) and other transformative technologies.

By harnessing AI-era technology, governments can reimagine climate action and deliver improved climate mitigation and adaptation that is more inclusive, efficient and effective to achieve real progress.

The Opportunity

Building upon categories identified by Google DeepMind, there are three ways AI-era technology will be critical in the fight against climate change: improving knowledge, optimising complex systems and accelerating breakthroughs.

Improving Knowledge of Climate and Weather

AI-era technology can help governments understand the world as it is – and as it is developing – so they can make better and faster decisions and predict future changes. For instance, digital twins can help solve discrete and connected problems with infrastructure such as electricity grids through experimentation. Tools like this depict not just present scenarios, but also past ones and a range of future possibilities to enable more effective planning and integration compared with traditional tools.

New tools also mean that governments can save lives, livelihoods and economies by anticipating crises and responding to them effectively. Companies such as Planet, which is partnering with the Tony Blair Institute (TBI), collect large data sets from satellite images to bring geospatial imagery to climate-vulnerable governments. These efforts can help to illuminate climate, conservation and humanitarian risks and realities as never before. Continual earth observation tracks land use day to day, helping efforts to detect deforestation, monitor changes in soil conditions and crop health, and coordinate humanitarian relief when disaster strikes.

In combination with AI and predictive analytics, satellite imagery can become truly transformative and inform a new generation of financial instruments, sustainable-management practices and innovative policy mechanisms. For example, weather-prediction tools that are underpinned by geospatial data – such as Google DeepMind’s GraphCast – recently outperformed conventional forecasting methods for the first time. Pairing this improved weather forecasting with cash transfers has enabled countries such as Ethiopia to increase their financial resilience in the wake of disasters by establishing shock-responsive social-safety-net programmes.

Optimising and Integrating Complex Systems More Easily

AI will be crucial in helping manage and optimise the systems and assets that are necessary to deliver net zero. As it understands and predicts processes significantly faster than humans can – and even in many instances understands processes that would be impossible for a person – AI can be used to improve how human systems operate.

For instance, the transition to renewable energy involves moving from a relatively simple centralised energy system to a complex one with intermittent and distributed generation, and flexible demand. Integrating this system and ensuring it works as effectively and efficiently as possible is paramount but complicated. With the use of AI, integration can happen more quickly, easily and securely.

There are already clear examples of how AI is being used to predict and optimise energy supply and demand patterns across electricity networks in real time. For instance, the company KrakenFlex hosts an intelligent demand platform called Smart Flex that integrates with energy devices, such as heating systems and electric-car charging units, in homes. The platform then coordinates the time when energy is used with the time when energy is cheapest and greenest. This improves reliability, helps manage the grid and results in huge efficiency and cost savings.

Accelerating Scientific Breakthroughs

The power of AI can massively reduce the experimental time needed to explore and identify requirements for new solutions to climate change. This opens significant potential for innovations – including the discovery of new materials – that can make climate action easier and cheaper. Through recently developed AI tools, researchers can rapidly find new material properties. These include battery designs that do not rely on rare-earth minerals, materials that capture CO2, chemistries that can produce carbon-neutral fuels, and replacements for materials that currently have few low-carbon alternatives such as steel and cement. For example, researchers recently discovered 2.2 million crystal structures – a finding that accelerates progress in fields from solar cells to superconductors and is equivalent to almost 800 years of experimentation.

AI can also help optimise food production by advancing knowledge of crop breeding. For instance, AI-powered tools can assist in identifying which genes could be edited to make crops more resilient and productive. The Gene Ranking Artificial Intelligence Network (GRAIN) platform uses an algorithm that evaluates scientific databases such as GenBank and identifies genes that act at a fundamental level to protect crops.

Actions Governments Can Take to Harness the Opportunity of AI

The gains that could be made by leveraging AI-era technologies are significant. Governments and global leaders need to fully grasp the potential that these technologies provide and put them at the centre of strategies for climate change, energy and infrastructure.

To do this, governments will need to make smart investments domestically and collaborate globally. This will involve taking six key actions:

  • Developing effective public-private partnerships: Most AI-era technology solutions are being developed by the private sector. Governments need to partner with these innovators to ensure that solutions are adopted rapidly. Initiatives such as Breakthrough Energy Ventures Europe, which was set up in 2019 by the European Commission, European Investment Bank and Breakthrough Energy Ventures, is an example of how governments and the private sector can work together to support innovation and bring new technologies to market. Governments should also support the rapid development of public-use cases through a permissive approach to private-sector technological experimentation, for example through regulatory sandboxes and tax incentives such as research-and-development (R&D) tax credits.

  • Treating data as public assets: Data relating to energy and the natural world should be considered key public assets. Governments should take the lead in creating highly valuable public-good data sets that can be used to train AI models for innovation and implementation of new solutions. Examples of such data sets include temperature data, road traffic and generation capacity on the electricity grid. For these data sets to have maximum value, it is crucial that they are accessible in standard formats to allow data to be efficiently merged from different sources. Where data sets are privacy restricted – for instance, in relation to household-energy demand profiles – there are opportunities to create synthetic and privacy-preserving versions of the data sets. For example, the Centre for Net Zero is doing this through its Faraday model, which creates demand profiles based on Octopus Energy customers’ smart-meter data in the United Kingdom.

  • Investing in compute infrastructure: Alongside talent and data, access to compute infrastructure has been cited by leading labs as one of the three key drivers of AI progress. Significant investment in compute infrastructure will enable training for and deployment of the most capable AI models. Critically, this infrastructure should be easy to upgrade and augment in a modular way over time – yet no country has a targeted plan to develop national AI compute capacity. To capture the potential of AI, governments need to create these plans and develop forward-looking policies that incentivise private-sector investment in compute infrastructure. Through smart and strategic choices, governments can mitigate new digital divides between countries with highly developed systems and those with emerging digital ecosystems (as highlighted in TBIs recent report, State of Compute Access: How to Bridge the New Digital Divide). Because compute infrastructure requires significant water and energy, it is also essential that governments implement policies and regulations to help drive efficiency. To do this, governments must make the most of emerging technologies that can increase server efficiency, such as free-air cooling and new machine-learning research, and also develop enabling policies for innovative approaches, such as using the heat from data centres to power nearby homes.

  • Aligning policy and regulation with AI-era technologies: In many cases AI-era technology has transformative impact, but the existing regulatory or policy regime does not incentivise desired outcomes. For instance, while AI could optimise flight paths to reduce the time that planes spend in the air, in most systems it takes years to approve new routes. Across key sectors, governments must update and modernise their regulatory and policy frameworks to ensure they enable, rather than block, the ability to use AI-era technology to achieve progress.

  • Developing cross-border learning, R&D and market development: Neither the policies nor technologies needed to fight climate change can be developed in silos. Economic openness is essential to drive climate action, and international cooperation is vital to create engagement across national borders. Governments must support research and lesson-sharing arrangements, develop cross-border markets and avoid the fragmentation of data, trade and ideas wherever possible. The fastest way to net zero is through collaboration and open competition, not through protectionism and isolation.

  • Integrating AI-era technologies into project delivery: The energy transition offers an opportunity for emerging markets and developing countries to build modern AI-era infrastructure upfront, and thus level up more quickly. Governments need to plan and invest in projects that harness AI-era technologies by default. From optimising hydropower operations to digitalising grids, the integration of AI-era technologies unlocks new routes to net zero and builds infrastructure that will be fit for purpose for decades to come. To do this, governments need to link planning, project preparation, procurement and inside-the-government-machine political work to drive through the policies and reforms needed to take projects from idea to investment.

Conclusion

Through harnessing the power of AI-era technologies, governments can enable a faster, better, more inclusive and cost-efficient route to climate action. AI can help accelerate the energy transition, improve products and services, reduce costs, and help states reap the economic benefits of the transition. Now is the time for leaders to act.

Newsletter

Practical Solutions
Radical Ideas
Practical Solutions
Radical Ideas
Practical Solutions
Radical Ideas
Practical Solutions
Radical Ideas
Radical Ideas
Practical Solutions
Radical Ideas
Practical Solutions
Radical Ideas
Practical Solutions
Radical Ideas
Practical Solutions