Is AI Computational Power the Currency of the Future?

4/1/24

Editorial team at Bits with Brains

Reflecting on the comments by OpenAI’s Sam Altman that computational power (“compute”) will be the future's currency, it's evident that AI's pervasiveness in our lives will soon make it a cornerstone of economic power.

The rapid commoditization of AI software and significant investments in AI chip manufacturing suggest that compute power will be a key driver of the future economy. As AI models become more accessible, the strategic importance will shift towards owning and creating the hardware and systems that power these models.

Nvidia's exponential growth, driven by its dominance in the GPU market crucial for AI computations, highlights the need to invest in chip manufacturing to maintain competitive margins as AI software standardizes. Nvidia's GPUs have become the de facto standard for training and running large AI models, giving the company a significant advantage in the AI hardware market. Other major players like Intel and AMD are also heavily investing in AI-optimized chips to capture their share.

Beyond hardware, the infrastructure supporting AI, such as data centers and electricity supply, is equally critical. Microsoft and OpenAI's $100 billion Project Stargate data center exemplifies the immense scale of investment required to stay at the forefront of AI development. This project aims to build a dedicated data center to train and run advanced AI models, which could provide Microsoft and OpenAI with a significant competitive advantage. Similarly, Meta is buying 350,000 Nvidia H100 GPUs.

This race to build powerful AI systems has led to unprecedented demand for electricity and advanced chip technology. Training and running large AI models requires enormous amounts of energy, which could strain electrical grids and contribute to carbon emissions if not managed sustainably. The future may involve pairing data centers with dedicated energy sources like nuclear power plants to meet these demands, redefining energy strategies for tech companies. This could lead to new business models and partnerships between technology firms and energy providers.

OpenAI CEO Sam Altman believes compute will be "maybe the most precious commodity in the world" and has suggested that computing power may become the world’s new reserve currency. The idea of using compute power as a global currency is intriguing but faces significant challenges. Supporters argue that in an AI-driven economy, controlling state-of-the-art compute infrastructure could be more strategically vital than traditional economic resources.

If compute serves as a de facto currency, it could spur greater investment from nations and companies in building advanced chips, data centers, and energy sources to power AI systems. In theory, a compute-based currency could also facilitate more efficient allocation of computational resources globally based on supply and demand dynamics.

However, there are numerous drawbacks to this concept. Technological barriers around standardization, verification, storage, and exchange of computational resources across disparate systems and parties would need to be overcome. The capital-intensive nature of building cutting-edge compute infrastructure could lead to market concentration among a few dominant tech giants and nations, raising concerns around centralization of power.

The energy-intensive nature of modern AI systems could incentivize unsustainable energy consumption unless paired with major advances in efficiency and renewable sources. We are already seeing some signs of this. The notion of compute overtaking traditional fiat currencies could also be perceived as an economic threat by nations and spark geopolitical conflicts.

A more plausible near-term scenario is compute emerging as an important store of value and medium of exchange within the technology sector, similar to digital assets like cryptocurrencies. Tech companies may increasingly use compute resources as a means of exchange or investment, creating a new asset class. However, entirely replacing traditional currencies seems improbable without extraordinary technological breakthroughs and international coordination.

The increasing importance of AI compute power has significant implications for businesses and policymakers. Senior executives should consider strategic investments in AI hardware and infrastructure to maintain a competitive edge. This could involve partnering with chip manufacturers or cloud providers to secure access to cutting-edge compute resources. Companies may also need to invest in talent with expertise in AI hardware and infrastructure to guide these strategic decisions.

Internally, businesses should prioritize AI initiatives that align with their core competencies and have the potential to create significant value. This may require reallocating resources and updating business models to center around AI. Executives should stay informed about emerging AI applications in their industry to identify opportunities and threats.

However, the capital intensity of advanced AI also raises policy concerns around industry concentration and regulation. If a few dominant players control the majority of AI compute power, it could stifle competition and innovation. Policymakers will need to carefully navigate these dynamics to promote a healthy AI ecosystem while mitigating risks of centralization and inequitable access. This may involve antitrust regulations, investments in public AI infrastructure, or incentives for smaller players to participate.

The energy demands of AI also necessitate a focus on sustainable solutions. Companies should explore innovative approaches to powering their AI operations, such as co-locating data centers with renewable energy sources or investing in more efficient cooling systems. Governments may need to incentivize sustainable AI development through targeted policies, such as carbon taxes or renewable energy subsidies. Collaboration between the tech industry and energy sector could lead to new solutions that balance performance and sustainability.

While compute power is unlikely to fully replace traditional currencies in the near term, its growing economic importance cannot be ignored. Businesses and nations that strategically invest in AI hardware, infrastructure, and sustainable energy solutions will be best positioned to capitalize on the transformative potential of artificial intelligence. This may require rethinking traditional economic models and measures of value to account for the rising importance of compute.

Proactive planning and adaptation will be key to thriving in an AI-driven future. Senior executives should prioritize staying informed about AI developments, investing in strategic capabilities, and fostering a culture of innovation. At the same time, policymakers will need to strike a balance between promoting AI innovation and mitigating risks around centralization, sustainability, and equity. International cooperation will also be critical to address the global implications of AI and ensure its benefits are widely shared.

Sources:

[1] https://ppl-ai-file-upload.s3.amazonaws.com/web/direct-files/1357227/bf75b11b-6dd6-4be8-85d8-8b27fa5ef419/Future currency of.pdf

[2] https://hbr.org/2023/08/what-the-finance-industry-tells-us-about-the-future-of-ai

[3] https://redresscompliance.com/a-guide-to-ai-in-finance-trends-and-future-predictions/

[4] https://www.datagrail.io/blog/data-privacy/generative-ai-privacy-issues/

[5] https://ovic.vic.gov.au/privacy/resources-for-organisations/artificial-intelligence-and-privacy-issues-and-challenges/

[6] https://www.worldfinance.com/markets/the-future-role-of-ai-in-finance

[7] https://economictimes.com/news/how-to/ai-and-privacy-the-privacy-concerns-surrounding-ai-its-potential-impact-on-personal-data/articleshow/99738234.cms

[8] https://link.springer.com/article/10.1007/s00146-021-01249-2

[9] https://www.thedigitalspeaker.com/privacy-age-ai-risks-challenges-solutions/

[10] https://www.neweratech.com/us/blog/role-of-artificial-intelligence-business-digital-transformation/

[11] https://www.isaca.org/resources/news-and-trends/isaca-now-blog/2021/beware-the-privacy-violations-in-artificial-intelligence-applications