High-profile tech leaders claim AI can help solve climate change, but critics argue that current generative AI, like ChatGPT, is contributing to rising emissions instead. While some AI applications are already improving climate research, doubts remain about the broader implications of AI’s energy consumption and effectiveness in practical climate solutions.
In an era where artificial intelligence is touted as the answer to pressing global issues, a lingering question emerges: Can AI like ChatGPT actually help solve climate change? Some proponents, including OpenAI CEO Sam Altman and Anthropic CEO Dario Amodei, believe these powerful technologies can be the solution. However, critics argue that the kind of AI contributing to rising greenhouse gas emissions isn’t the same type that will mend the climate crisis.
Alex de Vries, founder of Digiconomist, expressed skepticism about the optimistic view held by AI leaders. He noted, “I definitely wouldn’t count on these models to solve the problems of the world at the moment.” This skepticism surfaces against a backdrop of claims from influential voices in AI. At a recent event, former Google CEO Eric Schmidt suggested that society isn’t organized enough to reach climate goals on its own and prefers to bet on AI’s potential to make a difference.
In this same vein, Altman believes we’re only “a few thousand days away” from an AI superintelligence that will not only transform society but fix the climate as well. Amodei echoed this sentiment, envisaging powerful AI solving problems – but did mention that different technologies might lead to these advancements.
Yet, critics highlight that expecting future AI developments to offset current environmental harms is risky and speculative. While they acknowledge the urgency of climate change, they argue that existing technologies like solar panels and electric vehicles could drastically reduce emissions without waiting for AI to catch up. It’s all about the political will to implement change.
Interestingly, while AI-driven emissions are on the rise, the energy demands for training and running large-language models are staggering. Data centers, essential for AI operations, now consume about 3.7% of U.S. energy usage – and that figure is expected to grow. Reports from consultancy McKinsey predict this could rise to 11.7% by 2030. Meanwhile, worldwide energy consumption of data centers might double by that same year, putting pressure on global energy demands.
Despite the concerns, some AI forms are actively contributing to climate solutions. Studies have shown that machine learning aids in climate modeling and energy distribution management. Yet, these AI tools differ notably from the large-language models like ChatGPT, which critics claim are less energy-efficient. As Jathan Sadowski from Monash University points out, there’s confusion between generative AI and other machine-learning applications that are truly energy-efficient.
The path forward seems murky. Figures like Gary Marcus, an emeritus professor at NYU, remain cautious. With questions about reliability and practical utility, he stated, “Chatbots are not reliable enough even to follow the rules of chess, let alone make some stunning new advance in material science or something like that.”
When pressed for comments, Altman’s team redirected to prior statements, while representatives for Schmidt and Anthropic were silent on the matter. Delegates from the AI world seem to hinge much of their faith on future advancements — suggesting that generative AI will develop to levels surpassing human intelligence and creativity.
But experts like Emily Bender and Alex Hanna argue this view oversimplifies the technology’s capabilities, noting that AI lacks true understanding of the physical world. ChatGPT essentially predicts the next word in a sentence based on prior data — making it more of a fancy parlor trick than a viable solution to climate issues. In a bold comparison, Bender ruminated, “Are we going to have a Magic 8 ball solving climate change for us? I don’t think so.”
The debate surrounding AI’s role in climate change is complex and layered. While some industry leaders like Altman and Schmidt advocate for AI’s transformative potentials, critics emphasize the energy costs and current technologies that could alleviate climate issues without waiting for hypothetical advancements. The future remains uncertain, as the disparity between generative AI and more effective technologies raises questions. Without clearer paths to tangible outcomes, reliance on AI might be more wishful thinking than realistic.
Original Source: www.sfexaminer.com
