Post by account_disabled on Feb 17, 2024 2:08:16 GMT -7
Created by OpenAI, an artificial intelligence and research company, ChatGPT has caused a stir since its first months, with more than one million users in the first five days after its launch. This artificial intelligence tool allows you to have human-like conversations with a chatbot, offering potential benefits for humanity, but it has also raised some concerns.
Among the controversies surrounding its Middle East Mobile Number List use are legal and social responsibility issues. However, at a time when global collaboration is focused on limiting the planet's temperature, it is relevant to know the environmental footprint of ChatGPT and other similar models, as reported by The Conversation .
AI growth costs the planet
Chat Generative Pre-trained Transformer (ChatGPT) is a natural language processing tool powered by artificial intelligence technology. This language model, which is still in the research and feedback phase, can help answer questions and support tasks such as writing essays, emails, and even code.
However, Kate Saenko, associate professor of Computer Science at Boston University, mentions that the more powerful Artificial Intelligence (AI) is, the more energy it consumes. This implies that as generative AI models become more powerful, their power demand also increases. This raises concerns about society's future carbon footprint.
The emergence of more advanced generative AI models means that much more complex output can be created, such as sentences, paragraphs, images and even videos. Previously, generative AI was used in simpler applications, such as smart speakers or text autocompletion. However, today, these models have the ability to generate human language and create realistic images.
These studies are important for researchers, companies and decision makers to address these challenges and seek sustainable solutions that involve strategies to reduce the environmental footprint of ChatGPT and other similar models.
ChatGPT environmental footprint
Impact of ChatGPT's environmental footprint
However, it is currently difficult to accurately estimate ChatGPT's environmental footprint, which includes the energy used to manufacture the computing equipment, create the model, and use it in production.
In 2019, creating a generative AI model called BERT with 110 million parameters was found to consume the energy equivalent of a round-trip transcontinental flight for one person. The number of parameters refers to the size of the model, with larger models generally consuming more energy.
Based on these parameters, researchers have estimated that the creation of GPT-3, a much larger artificial intelligence model used as the technological basis to develop ChatGPT, with 175 billion parameters, consumed 1,287 megawatt hours of electricity and generated 552 tons of carbon dioxide equivalent. This is approximately equivalent to 123 gasoline-powered passenger vehicles driven over a year.
However, model size is not the only determining factor in carbon emissions. For example, the BLOOM model developed by the BigScience project in France is similar in size to GPT-3, but its carbon footprint is much smaller. This is due to the use of more efficient architectures and processors, as well as greener data centers.
Responsible use of AI
In addition to model size, a single generative AI query is also estimated to have a carbon footprint four to five times larger than a search engine query.
As chatbots and image generators become more popular and large companies like Google and Microsoft incorporate AI language models into their search engines, the number of queries generated daily is likely to increase exponentially and with it the impact environmental.
Another problem regarding the environmental footprint of ChatGPT and similar models is that they must be continually updated. For example, ChatGPT was only trained on data up to 2021, so it doesn't know anything about what happened since then. The carbon footprint of creating ChatGPT is not public information, but it is likely much higher than that of GPT-3. If it had to be recreated regularly to update its knowledge, the energy costs would increase even more.
Among the controversies surrounding its Middle East Mobile Number List use are legal and social responsibility issues. However, at a time when global collaboration is focused on limiting the planet's temperature, it is relevant to know the environmental footprint of ChatGPT and other similar models, as reported by The Conversation .
AI growth costs the planet
Chat Generative Pre-trained Transformer (ChatGPT) is a natural language processing tool powered by artificial intelligence technology. This language model, which is still in the research and feedback phase, can help answer questions and support tasks such as writing essays, emails, and even code.
However, Kate Saenko, associate professor of Computer Science at Boston University, mentions that the more powerful Artificial Intelligence (AI) is, the more energy it consumes. This implies that as generative AI models become more powerful, their power demand also increases. This raises concerns about society's future carbon footprint.
The emergence of more advanced generative AI models means that much more complex output can be created, such as sentences, paragraphs, images and even videos. Previously, generative AI was used in simpler applications, such as smart speakers or text autocompletion. However, today, these models have the ability to generate human language and create realistic images.
These studies are important for researchers, companies and decision makers to address these challenges and seek sustainable solutions that involve strategies to reduce the environmental footprint of ChatGPT and other similar models.
ChatGPT environmental footprint
Impact of ChatGPT's environmental footprint
However, it is currently difficult to accurately estimate ChatGPT's environmental footprint, which includes the energy used to manufacture the computing equipment, create the model, and use it in production.
In 2019, creating a generative AI model called BERT with 110 million parameters was found to consume the energy equivalent of a round-trip transcontinental flight for one person. The number of parameters refers to the size of the model, with larger models generally consuming more energy.
Based on these parameters, researchers have estimated that the creation of GPT-3, a much larger artificial intelligence model used as the technological basis to develop ChatGPT, with 175 billion parameters, consumed 1,287 megawatt hours of electricity and generated 552 tons of carbon dioxide equivalent. This is approximately equivalent to 123 gasoline-powered passenger vehicles driven over a year.
However, model size is not the only determining factor in carbon emissions. For example, the BLOOM model developed by the BigScience project in France is similar in size to GPT-3, but its carbon footprint is much smaller. This is due to the use of more efficient architectures and processors, as well as greener data centers.
Responsible use of AI
In addition to model size, a single generative AI query is also estimated to have a carbon footprint four to five times larger than a search engine query.
As chatbots and image generators become more popular and large companies like Google and Microsoft incorporate AI language models into their search engines, the number of queries generated daily is likely to increase exponentially and with it the impact environmental.
Another problem regarding the environmental footprint of ChatGPT and similar models is that they must be continually updated. For example, ChatGPT was only trained on data up to 2021, so it doesn't know anything about what happened since then. The carbon footprint of creating ChatGPT is not public information, but it is likely much higher than that of GPT-3. If it had to be recreated regularly to update its knowledge, the energy costs would increase even more.