Google’s Revelation
Data Centers, the physical locations where data is stored and managed, are critical to our digital activities. The article you're reading, social media, and AI tools are all supported by these supercomputers.
These AI tools are key players in the new industrial revolution, but they come with a significant downside. Google's recent environmental report revealed a 48% increase in its greenhouse gas emissions over the past five years. The culprit? Data Centers and the enormous energy consumption needed to train and operate AI systems.
These centers require vast amounts of water for cooling systems and even more electricity to keep them running 24/7 worldwide.
Google has brought this issue to light, but it affects all major tech companies, each striving to develop their own AI models in a relentless market race.
This leads to the pressing question: "Is the race for Artificial Intelligence Sustainable?"
The Energy Race
Adding to the complexity is the fact that AI is a rapidly evolving technology, becoming increasingly energy-intensive.
This means that the shift away from the stated goals of net-zero and carbon-neutrality for technology companies is likely to continue over the years or even amplify.
One strategy these tech giants have adopted is using low-cost, low-impact renewable energy sources. However, sustainable energy is currently scarce due to insufficient infrastructure investments.
The tech giants’ scramble for renewables could backfire: if they purchase these resources en masse without investing in new infrastructure, other companies will continue to rely on non-renewable energy, undermining global efforts to reduce carbon emissions.
Broader Implications
The gap between large tech companies and smaller entities can impact all aspects of sustainability, especially if we consider the broad spectrum covered by ESG (Environment, Social, and Governance) criteria.
Economically and socially, these large American corporations already have significant financial resources, allowing them to invest heavily in infrastructure, research and development, and talent acquisition.
This creates a competitive advantage but also places enormous responsibility and power in terms of privacy, cybersecurity, job process requalification, and managing biases and discrimination within AI learning models.
This concentration of power could lead to a near-monopoly on a global scale, potentially increasing economic inequalities between small and large enterprises and across different regions.
In countries like Italy, where 99% of businesses are SMEs, there's a risk of becoming overly dependent on tech giants’ AI platforms and cloud services, raising supplier policy risks.
Worse still is the potential for self-sabotage, using AI as an end in itself and wasting investments without a strategic vision.
The illusion of innovation often drives the mindset, “if everyone else is doing it, so should we.”
Adding further complexity to corporate governance are even the increasingly stringent EU regulations and policies on technology.
Two Paths to a Sustainable Future
These reflections are not from one who want to stop AI research. On the contrary, the potential of this technology is too significant to be ignored or delayed.
We must leverage it correctly, for instance, using its computational power to find new solutions to climate and social problems.
The goal of this article is to start a debate:
The first option favors large corporations, accelerating AI research but with evident short-term negative impacts.
The alternative requires greater international community effort, accurate infrastructure spending, global partnerships with developing countries to ensure no one is left behind, and investments in talent development.
These two paths represent crucial choices for the future: on one side, rapid progress hoping innovation will lead us to a sustainable world; on the other, an approach that embraces sustainability from the start, but requires global cooperation and long-term strategy.
.svg)
.svg)
.svg)