Artificial Intelligence (AI) has become a cornerstone of modern technology, influencing various aspects of our daily lives. However, the significant energy transfers associated with AI raise concerns about its environmental impact. For instance, training a single AI model can emit over 283,000 kilograms of carbon dioxide, nearly five times the lifetime emissions of the average car.
One reason for AI's high energy consumption is the need to synthesise new information. Unlike traditional search engines that retrieve existing data, AI models generate responses by processing vast amounts of data through complex computations. This process is akin to firing up a power station just to use a calculator, highlighting the substantial energy required for AI operations. This is the reason that data centres are so power hungry and the companies that run them are looking at developing small nuclear power stations to run them.
In contrast, traditional search engines are more energy-efficient. Their algorithms, honed over decades, are optimised to deliver search results using less computational power. For example, a standard Google search without AI integration consumes approximately 0.3 watt-hours (Wh) of electricity, equivalent to 1,080 joules, whereas an AI-driven search can consume about 2.9 Wh, or 10,440 joules. Human brains run at about around 20 Watts, it has been found that this increases when we re deep thinking or synthesising new information but I can't find reliable information as to what this amounts to, however there is some evidence that the higher your IQ the more energy transferred when trying to solve difficult problems!
To put this into perspective, consider the following analogy:
- Thinking for Yourself: This is akin to walking to your destination. It requires personal effort but involves minimal external energy transfers, reflecting the low energy usage of human thought processes.
- Traditional Search Engine Use: Comparable to taking the bus. It utilises more energy than walking but is relatively efficient, as the energy cost is distributed among many users.
- AI-Powered Search: Similar to driving a car alone. It offers convenience and speed but at a higher energy cost per user, leading to greater environmental impact.
This comparison highlights the varying energy demands of different information-gathering methods. While AI-powered searches provide advanced capabilities, they come with a significant increase in energy transfers compared to traditional methods. As we continue to integrate AI into various applications, it'scrucial to consider these environmental implications and strive for more sustainable practices.
Add comment
Comments