Unveiling the Hidden Climate Costs of Everyday AI: What You Need to Know

In 2014, Marissa Loewen began using artificial intelligence (AI) to manage her projects. As someone with autism and ADHD, she found AI helpful for organizing her thoughts. But, she’s conscious of the environmental impact of AI.

AI is everywhere now. It’s in our smartphones and search engines. Every time we use AI, it demands energy, often from fossil fuels, which releases greenhouse gases. This is a growing concern as we become more reliant on technology.

The Environmental Toll of AI

AI relies heavily on data centers, which process requests and store information. According to experts, the demand for these centers is rising. Noman Bashir, who studies computing and climate impact at MIT, states that many new data centers depend on fossil fuels because we can’t integrate enough renewable energy quickly enough.

These centers need cooling, consuming massive amounts of water. An article from the Environmental and Energy Study Institute notes that some large data centers can use up to 5 million gallons of water daily. This is similar to the daily needs of a town of 50,000.

Sasha Luccioni, an AI and climate expert from Hugging Face, points out that the impact of AI isn’t always clear. For example, generating a high-definition image uses as much energy as charging half a smartphone. Many users are surprised to learn just how much energy is consumed behind the scenes.

Professor Jon Ippolito from the University of Maine emphasizes that, while tech companies are working to make systems more efficient, AI’s environmental impact may not decrease. This is due to the Jevons Paradox, which suggests that as the use of a resource becomes cheaper, our consumption generally increases.

Understanding AI’s Carbon Footprint

The impact of AI on global warming is complex. It’s influenced by many factors, such as the local temperature and the energy source of the grid. Information on AI’s environmental costs can be hard to pin down, as data varies widely.

Ippolito created an app comparing the energy footprint of different digital tasks. He found that a simple AI query uses 23 times more energy than using Google without AI. For more complex prompts, energy use skyrockets.

Ironically, engaging in digital activities we take for granted often has a bigger environmental impact. For instance, an hour of Netflix consumes more energy than a complex AI task, and a group video call uses even more power.

Reducing Technology’s Climate Impact

To limit their environmental footprint, some experts advise reducing AI usage when possible. Ippolito suggests opting for human-made visuals instead of AI-generated ones. He also encourages users to keep questions simple and avoid unnecessary AI features.

Loewen practices similar habits. She aims to consolidate her queries to reduce energy consumption. She even built her own localized AI to minimize reliance on large data centers.

For a greener alternative, Luccioni uses Ecosia, a search engine that plants trees with its profits. This platform offers efficient algorithms and the option to turn off AI functions.

Interestingly, it’s worth noting that AI accounts for only a small part of data center energy use. Many resources are consumed by activities like social media and cryptocurrency mining. Ippolito recommends limiting social media time through screen time restrictions, which can reduce energy and data consumption.

Conclusion

It’s essential to understand the broader picture of technology’s environmental effects. While AI has its drawbacks, being mindful of all our digital activities can help lessen our impact on the planet.

For further insights, check out the Environmental and Energy Study Institute for more detailed statistics on energy consumption.

Read about the energy use of AI here.

Source link

Sasha Luccioni,Climate,Jon Ippolito,Marissa Loewen,environment,Business,Technology