Can quantum computing replace AI with a smaller carbon footprint?

With tech companies like Anthropic and OpenAI looking to file their Initial Public Offerings (IPOs), artificial intelligence (AI) is set to dominate the news cycle for a few more weeks. The hype around the stock price of AI companies will drown out the the issues of the massive carbon footprint of the data centers that power them. 

In such a scenario, the question is whether the next leap in technology i.e., quantum computing, helps us avoid the disasters caused by a major carbon footprint. 
Isn’t quantum power hungry too? 
Quantum computers rely on special quantum states to achieve superlative computing speeds. However, these quantum states can be reached only at temperatures near absolute zero, with a series of dilution refrigerators that consume large amounts of energy. 

A full-scale quantum computer ready for real-world deployment could easily consume 1 megawatt (MW) of energy, about what a small data center uses. So, where is the energy saving and small footprint? 

The answer lies in the computing capacity of these futuristic devices, which have been shown to complete calculations in a matter of minutes that even the fastest supercomputers today would take years to finish. At these capacities, global computing demand in 2040 could be met with fewer than 1,500 quantum computers. 

In comparison, meeting today’s AI and other computing demand requires more than 11,000 data centers, a Boston Consulting Group (BCG) study said. 
How will quantum computers help? 
Even when fault-tolerant quantum computers are ready to serve the world, it is unlikely they will be deployed as computing resources for chatbots, like chips are doing for AI today. 

Quantum computers will instead be deployed where they can create maximum impact, and that will be to solve problems of drug discovery, materials research, and climate change, and also affect carbon emissions through them.

The BCG study assessed over 50 applications where quantum computing could be deployed but identified only 16 where it could lead to a meaningful reduction of carbon emissions due to quantum advantage. 

These include building batteries with better electrolytes and electrodes that will store more energy, are environmentally friendly, and also cost-effective. This could help shift long-haul transport away from fossil fuels, thereby reducing emissions. 

Quantum computing will also help solve problems like direct air capture, where carbon dioxide is removed directly from the atmosphere, decreasing the concentration of the planet-warming gas. Another area is improving ammonia production processes, which are critical for food security but also harm the planet, contributing 1-2 percent of global emissions. 

There are some limitations to these projections, but if quantum computers can solve these problems, they will surely do it at a fraction of the energy and water consumption that today’s AI is doing.