On February 9, 2024, the Wall Street Journal reported that OpenAI CEO Sam Altman wants to raise up to $7 trillion for an extreme technology project to expand the global chip capacity and better power AI models. The CEO’s initiative to boost chip production could have significant environmental impacts, involving large amounts of natural resources, including water and rare earth minerals. AI tools’ water consumption has spiked and GPU shortages have become a major concern. In addition to its environmental impacts and chip war implications, OpenAI’s approach to AI development has faced criticism for its lack of efficiency.
The desire to address GPU shortages is not unique. Meta is also heavily investing in developing its own custom silicon for its AI workloads. Currently, there is limited transparency on the environmental impact of AI, but those facts do not look promising for the environment, considering Altman’s and others’ ambitious AI and silicon chip projects. AI GPU chips are cooled with water, and Fortune magazine has reported a 34% increase in Microsoft’s water usage due to its 2023 AI tools. Additionally, Meta’s Llama 2 AI model guzzled twice as much water as did its Llama 1, and OpenAI’s GPT-3 training alone consumed about 185,000 gallons of water.
The whytry.ai article you just read is a brief synopsis; the original article can be found here: Read the Full Article…
