And in case that news was not daunting enough, a new controversial study is raising plenty of eyebrows across the board linked to how AI might be consuming way more energy than anybody may have ever imagined.
Yes, we’re talking about those generative AI models that continue to be debated in terms of how beneficial they really are for the human race. But what many people may have overlooked is how big of a carbon footprint they emit and how much energy they use.
This week, a team of top research experts sat down to look at the matter closely. According to them, AI systems that use ChatGPT manage to roll out close to 130 to 1500 times less carbon when compared to humans.
The same goes for producing pictures and they took the latest DALL-E-2 into consideration and rolled out nearly 310 to 2900 times less carbon.
This paper then came up with the conclusion that using AI has the ability to accomplish a few major activities including much less carbon emissions than when humans are at work. Now it all sounds like a picture-perfect ending, but wait, there’s a lot more to the story than just that.
Not everyone is happy and more dialogue is necessary, experts claim who feel the study needs to be rechecked because the challenges that such technology poses in terms of devastating effects on the climate may be overlooked.
On the other hand, we had one top scientist put down his two cents on the matter. He alongside experts from MIT and the University of California added how the matter is not so simple. And to get a better understanding of what’s going on, both the world of AI and Blockchain needs to be looked further into detail.
Today, there is plenty of existing data that proves the environmental effects that AI systems have and how they can impact the environment. The text compares yearly carbon footprints produced by humans to those produced by AI models generating images and text.
And that’s when researchers signaled the significance of taking carbon emissions into consideration from other leading activities such as AI to ensure the right type of policies are in place regarding sustainability.
This is why experts felt that authors cannot make any such claims about AI being less harmful to the environment and proving to be more sustainable. There is no ground information that proves this and that’s super necessary.
Man, this preprint is really the gift that keeps on giving.
— Sasha Luccioni, PhD 💻🌎🦋✨🤗 (@SashaMTL) September 19, 2023
In case people missed my previous PSA : you can't compare the carbon emissions of people and objects. Humans are more than just the work that they do.
(Also, that paper makes a lot of false assumptions in general) https://t.co/bZA414J9YI
Remember, using AI as the most creative means without considering its damaging effects and incredible harm will truly be super irresponsible. And if it does end up doing this much harm, it’s time people reevaluate whether or not it’s worth taking the leap into this field.
Some context was also put forward regarding the former analysis linked to the blockchain domain. Then another point worth discussing was rebound effects and how more efficiency led to greater usage.
Other experts that it was controversial to compare humans to objects and therefore the findings cannot be taken.
See, it’s all a very complex matter and something that cannot be avoided, no matter how hard you try. It’s a blend of climate, AI, and society. So yes, it’s an ongoing debate and one that will need more answers with evidence from literature studies to weigh out all the benefits from the drawbacks.
Read next: New Survey Says AI Implementation Sees A Surge In Investments With Positive Responses From Employees
by Dr. Hura Anwar via Digital Information World
No comments:
Post a Comment