CLIMATE change is considered the most critical challenge of our time, while artificial intelligence (AI) is the most advanced innovation of this generation. The question arises: What would pitting cutting-edge technologies against nature’s destructive forces look like?
This corner pitched the idea of something like Goliath vs. Godzilla in a raging battle to cut the other down to size. The added twist was to explore the upsides as well as the downsides of using AI to reduce the hazards and risks of escalating climate change.
IBM responded by arranging an interview with Arun Biswas, one of Big Blue’s Asia-Pacific leaders for strategic engagements under IBM Consulting. Biswas is also an industry innovator and an AI and sustainability advocate with more than two decades of experience in technology-enabled business transformation.
Here’s how our conversation went:
THE MANILA TIMES (TMT): AI is often perceived as a double-edged sword, with upsides and downsides in all spheres of human endeavor. What are the advantages of using AI to forestall or even prevent the adverse onset of climate change?
ARUN BISWAS (Biswas): Sure, sure. That’s a very important topic — something that is also very close to my heart.
We have to think about the impacts of AI on climate change in two parts. One is how AI helps mitigate climate change. The second is how AI helps communities adapt better to climate change.
Let me start with mitigation and adaptation. If you think about mitigation, there are a number of ways in which AI can help reduce emissions, which are the main cause of climate change.
First, AI is very good at helping integrate renewable energy into the energy system, such as solar and wind. The big problem with these renewable energy solutions is that they are intermittent. The sun does not always shine, the wind does not always blow, but the electricity grid demands a stable power source. In this context, AI is very good at balancing electricity supply and demand.
When you are integrating renewable energy sources — intermittent sources, at that — you need high-precision forecasts, and AI excels at energy forecasting. AI allows the grid to stay balanced at all times. It also helps in the generation of renewable energy. For example, in the case of solar energy, AI helps in site planning and in maintaining and operating solar assets in a more efficient and autonomous manner.
The second benefit is energy efficiency. AI can optimize energy use. If you are using a certain amount of energy in an industrial process, AI can be used to optimize power utilization. Think about energy savings in buildings and facilities.
The third way is measurement and monitoring. AI can help measure and monitor emissions more accurately by using satellite data, drone data and other sources, and subsequently help analyze this data with greater precision.
Large-scale innovation is the fourth. For example, AI can help discover materials for new batteries, new kinds of solar panels that are more efficient, or even materials that can absorb carbon more effectively. In that sense, AI helps cut emissions faster and with greater precision, essentially slowing climate change.
Next, let’s look at adaptation — how AI can prepare us for and manage climate impacts. The first is early warning systems. Climate change is largely felt in the Philippines, for example, through more typhoons, more frequent flooding and intense heat waves. In these challenges, AI can act as an early warning system to predict their onset earlier, giving people a chance to prepare for these events.
Second is climate risk mapping. AI can identify flood zones, areas vulnerable to sea-level rise and zones with high landslide risk. These are crucial in helping build resilient infrastructure. AI can analyze large volumes of satellite data and identify risky zones where infrastructure needs to be built differently.
The third area is agriculture. AI can help farmers optimize irrigation and adopt precision agriculture. This is especially important because, as the climate changes, many farmers find that traditional crops are no longer viable. They may need to switch crops or adjust water use, fertilizer application or pest control cycles. AI can help farmers in these situations.
A fourth area is health. Climate change is predicted to lead to new kinds of disease outbreaks, and AI can help predict climate-linked disease patterns. It can also address heat-related hazards and help people adapt to heat-risk patterns more effectively.
Finally, there is disaster recovery and response. Before a disaster hits and during the disaster itself, AI can assess damage quickly and accelerate restoration and relief planning.
TMT: It has been reported that developing AI itself has downsides. Can these disadvantages offset the gains against global warming?
Biswas: Yes, indeed, there are downsides to AI that we must be aware of. The first is the energy intensity of AI itself. AI models consume a lot of energy during both the training phase and the use phase, which we call inference. If this energy is supplied by fossil fuel–based power grids, that becomes problematic.
The second is water consumption. AI systems run on data centers, which require extensive cooling because AI chips generate heat. Data center cooling often relies on water, which can be especially problematic in water-stressed regions.
The third area is critical materials dependency. AI infrastructure requires materials such as lithium, cobalt, nickel, copper and rare earth elements. Mining these materials is carbon-intensive and can damage the environment and ecosystems.
The fourth potential downside is electronic waste. This can create problems, especially in countries where recycling capacity is limited.
So AI has real environmental and energy costs. But the good news is that the benefits far outweigh the costs, and many of these costs can be mitigated — even with AI itself — through energy-efficient design, sustainable data centers and circular material use. Overall, the benefits clearly outweigh the costs.
TMT: AI development is concentrated in developed countries. Does this create a dependency for developing countries when accessing AI capabilities to address climate change?
Biswas: That is the reality. At the same time, AI allows developing countries to leapfrog, and that is an opportunity they can seize. For example, energy systems in the Philippines can be made much smarter through AI, even without major upgrades to physical infrastructure.
Another way to look at it is that AI can reduce infrastructure costs by extending the lifespan of critical assets through AI-based predictive maintenance. When you do that, you reduce the need for capital-intensive investments in new infrastructure.