January 9, 2024
​

The Rebel Tech Newsletter is our safe place to critique data and tech algorithms, processes, and systems. We highlight a recent data article in the news and share resources to help you dig deeper in understand how our digital world operates. DataedX Group helps data educators, scholars and practitioners learn how to make responsible data connections. We help you source remedies and interventions based on the needs of your team or organization.

​

IN DATA NEWS

“A recent study by Hugging Face and Carnegie Mellon University found that using AI to perform various tasks such as generating images or writing emails incurs a significant energy and carbon cost. Generating one image using a powerful AI model can consume as much energy as charging a smartphone, while text generation is less energy-intensive. The findings suggest a need to reconsider the use of generative AI in favor of less carbon-intensive alternatives.”

It’s a new year and we’re still uncovering the depths of algorithmic-based harms. Earth 1 is being traumatized by AI in terms of water and energy consumption.

• Training ChatGPT-3 took an estimated 700,000 liters (or 185,000 gallons) of water.
• Training ChatGPT-4 took an estimated 3,500,000 liters (or 925,000 gallons) of water.
• 5 to 50 ChatGPT prompts use 500 milliliters (or 16 fluid ounces) of water.
• Training ChatGPT-4 took an estimated 1300 megawatts of energy. That’s the average energy consumption equivalent of 1,000 US households for 5 or 6 years.
• 1 ChatGPT prompt uses 15 times more energy than a Google search query.
• Generating 1 AI image consumes an average of 15 minutes of powering a typical refrigerator.
• Generating 1 AI image consumes as much energy as charging your smartphone.
• Generating 1,000 images emits the same carbon dioxide equivalent as driving 4.1 miles in a gas-powered car.

And it’s important to not list these numbers without sharing how they scale. There are approximately 10 million ChatGPT prompts processed daily with a cost of 7 cents per prompt. So, after doing some quick ChatGPT math, we have some astounding figures: costing $700,000 per day and using 100,000-1,00,000 liters of water per day. Images, as to be expected, cost more and use more water. Nearly 34 million AI-generated images are created daily. The energy needs therefore average 8.5 million hours of refrigerator powering time or fully charging 34 million smartphones.

So now that we’re learning some of the concrete costs to our natural resources based on current estimated levels of consumption, we need to choose responsibly and ethically when and how often to use AI tools, systems, and platforms. Our carbon footprints are casting a long, deep, and wide shadow we can’t conveniently overlook. Earth 1 needs all of us to reduce our collective use of AI.

​

​

Like what you're reading? Find it informative and insightful? You can sponsor the Rebel Tech Newsletter and follow on LinkedIn.

​

​

"Tech moves fast. The law moves slow. Coping with these two extremes causes a host of varying intensities of friction for people familiar with tech, for those attuned to the law, and for everyone else in between." pg 215 Data Conscience

​

​​​

​HAPPENINGS & APPEARANCES

• We’ve recently returned from our winter sabbatical. Feeling refreshed, renewed, and ready for what’s to come in 2024 for the data/AI ethics industry.
• Equitable AI for the electric power sector is a new conversation for DataedX. It’s encouraging to learn and engage with more forward-thinking organizations that are prioritizing the implementation of ethical data/AI practices and professional development.
• Dr. Marshall, representing TRAILS, has joined the SIG Ethics and Trustworthiness group’s leadership team.
• Dr. Marshall has joined a workgroup to inform the update of UPenn’s A Toolkit for Centering Racial Equity Throughout Data Integration.
• Take a look at Kapor Center’s Justice-Centered Computer Science Education Initiative framework, where Dr. Marshall served as a senior advisory board member.

​

LAUGHING IS GOOD FOR THE SOUL

​

Stay Rebel Techie,

Dr. Brandeis

​

Thanks for subscribing! If you like what you read or use it as a resource, please share the newsletter signup with three friends!

​