At the recent World Economic Forum held in Davos, Switzerland, considerable emphasis was placed on the future of artificial intelligence (AI), particularly in relation to energy consumption. OpenAI CEO Sam Altman emphasized the urgent need for an energy breakthrough to support the development and operation of advanced AI models such as ChatGPT and Claude from Anthropic. Altman’s call to action reflects growing concern in the tech industry about the environmental footprint of AI technologies.
The energy intensity of AI
The power requirements for AI, especially for large language models (LLM), are enormous. Altman pointed out that current methods for developing such AI models involve the use of tens of thousands of graphics processing units (GPUs), resulting in incredibly high power requirements both during the training and post-training cycles. The situation is such that the energy consumption of AI models may soon equal the energy consumption of an entire country.
The Fusion solution
In response to these challenges, Altman advocated investment in nuclear fusion as a potential solution. Fusion energy, which involves fusing atoms together to release energy, is seen as a zero-carbon solution that could scale electricity grids to meet the demands of AI technologies. Altman has already invested heavily in this area, contributing $375 million to Helion Energy, a fusion startup, in 2021. Since then, Helion has signed a deal to provide energy to Microsoft, which is not only the biggest financial backer of OpenAI, but also provides computing resources for AI development.
A global perspective
Speaking at a Bloomberg event on the sidelines of the WEF meeting, Altman highlighted the need for more climate-friendly energy sources, including cheaper solar and storage, along with fusion. He also expressed a desire for the world to accept nuclear fission as a source of energy. This position highlights the importance of exploring multiple avenues to meet the energy needs of future AI models.
The urgency of action
The call for an energy breakthrough is not just to enable the next generation of AI technologies; it is also about addressing the environmental impact of these achievements. The carbon footprint of the AI industry is becoming increasingly apparent, raising concerns about the sustainability of current practices. As artificial intelligence becomes more integrated into various sectors, its energy consumption will increase dramatically, requiring immediate action and innovation in energy production and storage.
Sam Altman’s remarks at the Davos forum highlight the critical intersection of AI development and environmental sustainability. As AI technologies advance, the industry must simultaneously address escalating energy demand and environmental impact. Pursuing solutions like fusion power not only promises to support the growth of AI, but also contributes to the broader goal of achieving a more sustainable and greener future.
Image source: Shutterstock