It’s official, in the last 12 months, the Earth’s average temperature has now exceeded pre-industrial levels by 1.5 degrees Celsius. 

Despite the 2016 Paris Climate Accord, a landmark international treaty where almost 200 nations pledged to hold the Earth’s temperatures below 2 degrees Celsius above pre-industrial levels, we have now come to a tipping point where governments have to seriously question whether they are realistically able to limit this increase.

This article was written in 2016: “30 years ago scientists warned Congress on global warming.” Now, almost 37 years later, after the US Senate on Environment and Public works held these crucial hearings on Ozone Depletion, the Greenhouse Effect, and Climate Change, the needle has failed to move in a substantive way.

The late Senator John H. Chafee of Rhode Island, who convened these hearings said this at the time: “This is not a matter of Chicken Little telling us the sky is falling,” Chafee said at the hearing. “The scientific evidence is telling us we have a problem, a serious problem.”

What’s clear is that currently, in 2024, humanity has done little to transition away from fossil fuel. It is incredibly hard to reduce the dependence on oil. There was a fleeting moment where the Coronavirus was a catalyst for substantial change. The world-wide stay-at-home orders created this plunging economic activity that included the plummeting demand for oil. It was indeed short-lived. This event slowed down oil production for a little while, but the Russian invasion of Ukraine created another panic to move away from Russian oil, which resulted in an escalating oil prices. I heard this on the Economist: Humanity has been living in the age of fossil fuels for more than a century. Transitioning away from a resource that has and continues to be effective and efficient is not easy.

As long as the oil and gas industry continues to generate significant profits, corporations will continue to produce fossil fuels. Economies that rely on oil will fuel this self-fulfilling reality. The political football that peddles job losses and a domino effect on communities that rely on this resource further slows this transition.

Oil remains plentiful. It is economical to produce and use. There has been a lack of aggressive campaigns towards alternative energy sources. What I’ve witnessed is that with new energy alternatives comes rethinking storage, distribution and infrastructure.  I spoke to experts including David Agbanwu to understand what this means as a deviation from the systems we know. Despite the promises of virtual power plants, the coordinated efforts among, wind, solar and battery towards electrification, the efforts to make this available everywhere, especially to remote communities and developing nations still highly dependent on diesel technology, is daunting. I concluded, “The possibilities of electrification across a global spectrum, continue to be, however, steeped in political controversy, investment, and supply chain challenges.”

‍

Technology’s Addiction to Power

I work in technology, and I have witnessed the evolution from AI/ML from narrow AI towards general AI and the ensuing amount of energy that has surged since the rise of ChatGPT. Sasha Luccioni, researcher and climate lead at HuggingFace, wrote about the “The mounting human and environmental costs of generative AI”. 

There was a time when the mining of cryptocurrency was responsible for excessive consumption of energy. Between 2016-2018 this study concludes that Bitcoin accounted for between 3-15 million tonnes of C02 emissions. Compared to the US total emissions of 5,311 million tonnes, this represents 0.19% of the US annual carbon emissions.  The example given in this study: “mining one bitcoin in India emits ~6400 kilograms of CO2.”

Let’s compare this to large language models (LLMs).  Companies pay based on the amount of data processed by the model. This is determined by the number of tokens (words, symbols) processed. According to Luccioni, AI models have gotten bigger. Measured by the number of parameters (or weights denoted from the patterns retrieved from trained data), LLMs like ChatGPT, this metric has soared from 100 million in 2018 to now over 500 million in 2023. The cost of training these large models is substantial. Only a few organizations the size of OpenAI can afford these: “…the training cost of GPT-3, which has 175 billion parameters, at $4.6 million.”

LLMs demand substantial computational resources and infrastructure investment. On a smaller scale, hosting an LLM on AWS (for example) costs nearly USD$33 per hour, totaling at least USD$23,000 per month.

The irony is that the very technology that is supposed to drive progress, make things more efficient, and make humans more productive, have significant energy and financial costs. Balancing energy with environmental impact remains a challenge.

Oil continues to serve an incredible value to keep these technologies running. This resource is essential to complement renewable energy sources to stabilize the grid.  In remote areas where connections to the larger grid remains a challenge, diesel and oil-based generators provide the essential electricity.

There is clearly a push to transition to cleaner fuels but at this juncture, the progress has been slow. 

The question to the institutions that are making substantial profits from technology and governments who are the purveyors of progress, and at the same time, are responsible for long term impacts to people and the planet: As we build towards a future of increasing convenience, productivity, and a better life for everyone, who are we building for, and for what purpose?  More importantly, whose values do we serve? The clock’s ticking and we do not have the luxury of another 30 years to do the right thing. 

‍

Hessie Jones, with her impressive background, stands out as a contributor to Forbes and a former Senior Marketing Manager at Yahoo!. Her leadership journeys, including roles as CEO of ArCompany and AVP of Internet Marketing at Citibank, underscore her expertise and influence in the field of marketing. Hessie's diverse experience and ability to write on a multitude of topics not only showcase her versatility but also offer readers a deep dive into the complexities of the technology and digital landscapes.

‍

‍