IBM Says Quantum Computing May Solve AI’s Energy and Cost Crisis

The exponential growth in AI training costs is raising sustainability and equity concerns, but quantum computing may offer a solution.

The AI training cost problem: The rapid increase in AI usage is driving up energy consumption and costs, which could surpass the US GDP by 2030 and limit AI’s potential if left unchecked.

• ChatGPT queries consume at least 10 times more energy than Google searches, contributing to the unsustainable growth in AI’s energy footprint.
• By 2030, the cost to train a single AI model once could exceed worldwide IT spending, creating a hard ceiling for the technology’s development and accessibility.

Corporate responsibility and mitigation efforts: Some companies are taking steps to address the environmental and financial challenges posed by AI’s growth:

• AWS is implementing liquid cooling solutions, exploring alternative fuels, and developing more efficient chips like Trainium and Inferentia to reduce carbon usage and improve performance per watt.
• AWS’s second-generation ultra cluster network supports up to 20,000 GPUs, delivering faster training at lower costs through increased network throughput and reduced latency.

Quantum computing as a potential solution: IBM is researching quantum machine learning’s potential to provide resource savings and speed benefits for AI:

• Quantum computers could offer advantages for tasks involving limited, sparse, or highly interconnected data, particularly in healthcare and life sciences applications.
• IBM’s Watson Code Assist aims to help users leverage quantum computing for their applications, even if they are unfamiliar with the technology.
• The future likely involves a combination of CPUs, GPUs, and quantum processors working together to solve problems efficiently, but further advancements in quantum infrastructure and physics research are needed.

Transparency and choice in AI adoption: Experts emphasize the importance of transparency and choice in enterprise AI adoption to ensure sustainability and equity:

• Decision-makers need a deep understanding of the sustainability, energy, privacy, and security characteristics of AI technologies to calculate the true return on investment.
• Organizations should carefully consider their use cases and choose the most cost-efficient and optimal deployment options, such as selecting appropriate silicon and hosting solutions.
• Pushing for transparency and exploring alternatives to monolithic models, such as multiple smaller models, can help mitigate the risks of hitting a hard ceiling in AI development.

Broader implications: The exponential growth in AI training costs highlights the urgent need for sustainable and equitable solutions to ensure the technology’s long-term viability and accessibility. While corporate efforts and quantum computing offer potential paths forward, a concerted effort from industry, researchers, and policymakers will be crucial in addressing this complex challenge. The decisions made now will have far-reaching consequences for the future of AI and its impact on society.