Industry
Artificial Intelligence has a bit of a bad rep at the moment. It feels like some of the early excitement has faded, replaced by a wave of AI anxiety with people worrying about jobs disappearing or the environmental impact of energy hungry data centres. But are we losing sight of how much we stand to gain from this technology? From new drug discovery to improved weather forecasting, and now breakthroughs in battery innovation, AI has the potential to tackle some of humanity’s biggest challenges. So is AI really the villain we’re making it out to be, or could it play a far more positive role in the years ahead?
Why better batteries matter for electrification
As concerns around climate change mount, all eyes are on electrification. Once regarded as a luxury, electric vehicles (EV) are now widespread. The growing popularity of EVs has been influenced by improved charging infrastructure and government initiatives, as well as innovations in battery technology, which have reduced their price dramatically. Take lithium iron phosphate batteries for example. They were once seen as unsuitable because of concerns around lower energy density. Focused research and development by Chinese producers has changed that and turned them into a practical alternative to the earlier chemistries they were measured against. Today they account for more than half of the global EV market and cost roughly 30% less than competing options.
Despite improvements in the efficiency and cost of lithium ion batteries, which make up virtually all EV batteries, there is growing pressure to transition to new, more abundant battery chemistries. This is because lithium mining and extraction, most prevalent in Australia, Chile, and China, poses huge environmental impacts. Research shows that around 2.2 million litres of water are needed to produce just one ton of lithium, and in one of Chile’s key mining regions, lithium and copper extraction has used over 65% of the local water supply, affecting both farmers and fragile ecosystems. EV production also has roughly double the carbon footprint of a standard internal combustion engine vehicle. If we can reduce our reliance on lithium and discover new battery chemistries that use more naturally abundant materials, we can cut the environmental cost of electrification.
How AI is accelerating battery innovation
Luckily, AI is accelerating battery innovation by allowing researchers to explore thousands of potential materials in record time and pinpoint the few that are worth testing. One promising new type of battery is the multivalent ion battery. Lithium ions carry a single positive charge, but metal ions like magnesium, aluminium, and calcium carry more. In theory, this could unlock batteries with much higher energy density made from materials that are far more available. But making these batteries a practical reality means finding structures that can hold and transport these ions effectively. In new research, teams at the New Jersey Institute of Technology used AI to sift through a pool of millions of potential structures, narrowing down thousands of options into a shortlist of promising candidates.
Another exciting direction is solid state batteries. You’ve probably seen stories about lithium ion batteries catching fire in waste trucks, on aeroplanes, or during e-bike charging. By replacing liquid electrolytes with solid materials, solid state batteries remove this fire risk and are not only a safer but faster charging alternative. A team at Microsoft Quantum recently used AI to screen over 32 million potential materials for a new solid state electrolyte, something that would take humans years. They narrowed the list to eighteen and eventually identified one promising candidate that uses 70% less lithium by incorporating sodium instead, and they even managed to power a lightbulb with it.
A reminder of what we stand to gain
AI might feel overwhelming at times, but it also holds the keys to scientific breakthroughs we have been trying to unlock for decades. In drug discovery, we need it to help us tackle antibiotic resistance and support rapid vaccine development, since microbes can mutate faster than we can respond. AI can even help fix some of the problems it currently creates. While everyone knows AI uses a lot of energy, it can also help us use less. According to the International Energy Agency, if AI is widely adopted, optimised production processes could deliver around eight percent energy savings in light industry by 2035, covering food, clothing, and electronics.
And this same potential is already showing up elsewhere. In the battery space, AI is helping researchers uncover new chemistries far more quickly than traditional methods allow. With quantum computing on the horizon and able to run complex simulations using far less compute power, the pace of discovery is only set to climb. All of this points to a simple truth. It might be time to look past the negative headlines and take a more balanced view of the role AI could play in the years ahead.
