Crypto & Blockchain News

The Environmental Impact of Artificial Intelligence: Concerns Beyond Misinformation and Job Threats

In the realm of artificial intelligence (AI), discussions often revolve around misinformation and potential threats to human jobs. However, a Boston University professor, Kate Saenko, drew attention to another important issue—the substantial environmental impact of generative AI tools.

As an AI researcher, Saenko raised concerns about the energy costs of creating AI models. In an article in The Conversation, he emphasized, “The stronger the AI, the more energy it requires.”

While the energy consumption of cryptocurrencies such as Bitcoin and Ethereum has generated widespread debate, the rapid development of AI has not received the same level of scrutiny when it comes to its impact on the planet.

Professor Saenko aims to change this narrative, acknowledging the limited data available on the carbon footprint of a single generative AI query. However, he highlighted that studies show that energy consumption is four to five times higher than a simple search engine query.

A landmark study from 2019 examined a generative AI model called Bidirectional Encoder Representations from Transformers (BERT), which consists of 110 million parameters. The model consumes the energy equivalent of a round-trip transcontinental flight for one person during its training process, using a graphics processing unit (GPU). Parameters, which guide model predictions and increase complexity, are adjusted during training to reduce errors.

In comparison, Saenko revealed that OpenAI’s GPT-3 model, with a staggering 175 billion parameters, consumes the energy equivalent of 123 gasoline-powered passenger vehicles driven for one year or about 1,287 megawatt hours of electricity. In addition, it produces a staggering 552 tons of carbon dioxide. Remarkably, this energy expenditure occurred before any consumer even started using the model.

With the growing popularity of AI chatbots, such as Perplexity AI and Microsoft’s ChatGPT integrated into Bing, the situation has been exacerbated with the release of mobile apps, making these technologies even more accessible to a wider audience.

Fortunately, Saenko highlights a study by Google which proposes various strategies to reduce carbon footprint. Adopting a more efficient green architecture of models, processors, and data centers can substantially reduce energy consumption.

While one big AI model may not destroy the environment alone, Saenko warns that if multiple companies develop slightly different AI bots for various purposes, each serving millions of customers, the cumulative energy usage could become a significant concern.

Ultimately, Saenko suggests that further research is essential to increase the efficiency of generative AI. With passion, he highlighted the potential of AI to operate on renewable energy sources. By optimizing calculations to coincide with the availability of green energy or locating data centers where renewable energy is abundant, emissions can be reduced by an extraordinary factor of 30 to 40 compared to relying on a grid dominated by fossil fuels.

In conclusion, while concerns about misinformation and job displacement due to AI remain, Professor Saenko’s emphasis on the environmental impact of generative AI tools poses a critical issue. This requires increased research and innovative approaches to ensure that AI development is in line with sustainability objectives. In doing so, we can harness the potential of AI while minimizing its carbon footprint, thereby paving the way for a greener future.

Source link

Related Articles

Back to top button