Contents
Artificial intelligence (AI) tools like ChatGPT are already replacing some copywriting jobs. Nevertheless, many people are not sitting at their desks waiting to be sacked by AI. Climate quitters have virally multiplied over the last few years. Plus, youngsters want to start their career with greener jobs. CEOs are also worried. Last November PwC surveyed 4,702 executives across the world. 45% of them said their business journey may be over in 10 years if they don’t do anything to address AI and climate impacts.
Based on this scenario, I thought of putting my zombie hat on and picking some human brains about how AI will change the climate change industry. I reached out to 5 experts and asked them for their educated guess on the following two questions:Â
Questions for the experts
How can AI help us solve climate change?
How will artificial intelligence affect jobs in the climate sector?
Is AI Bad For Climate Change?
Before hearing experts’ views on question 1, we should look at the environmental impact of AI, i.e. its carbon emissions. While many people are caught up in the AI excitement, questions about its true environmental cost still lurk behind the scenes.Â
First of all, just like humans, machines also need training. That’s where the machine learning term comes into play. Well, according to a recent study by the start-up Hugging Face, it took a very carbon-heavy lesson to train their model BLOOM. How heavy? Just Like charging 3 million smartphones!
So, why does artificial intelligence have such a significant carbon footprint? To understand this, we need to consider the energy-intensive nature of training these sophisticated models. The computational demands of performing this operation, often involving millions or billions of parameters, are immense. On top of that, AI training is often powered by fossil fuels. That’s not all. When taking into account the carbon emissions produced by manufacturing the trainer, a.k.a. huge computing infrastructure, figures double. As is the case with BLOOM.
Having said that, we should look at the other side of the coin as well. Some experts argue that, like any emerging technology, AI will become more environmentally friendly. For instance, Green AI is promoting more efficient data collection and processing, thus cutting the energy consumption of training machine learning models. On the other hand, the United Nations Environment Program (UNEP) praises AI as an innovative tool with great potential to help tackle environmental challenges such as detecting microplastic in wastewater. Also, AI is already helping businesses with sustainability. But it’s not just about the environment. Artificial intelligence should also be socially responsible!
What Is Responsible Artificial Intelligence?
The question 2 mentioned above aims to highlight one of the ethical implications of AI, that is, the impact of automation on human job security. So, it’s worth giving a brief introduction to the concept of responsible artificial intelligence.Â
Like apps, social media, or any other powerful technology, AI should be used responsibly. When talking about responsible AI, there are a series of principles (e.g., transparency, data privacy, etc.) to keep in mind.Â
When focusing on question 2, the most straightforward solution would be that companies help their workers up-skill themselves. For instance, copywriters can learn how to turn chatGPT into a user-friendly resource rather than seeing it as a job-stealing machine. Speaking of the robotic devil…
1. ChatGPT
While it might seem paradoxical to some of you, I asked my 2 initial questions to chatGPTÂ as well. Why? Because it allows us to have a non-humane baseline to start from. So, here you go!
Bio
ChatGPT was released on 30 November 2022 by OpenAI and its potential has grown exponentially since then. ChatGPT is proficient at responding to users’ prompts, writing any kind of deliverable (e.g., CV, cover letter, book, wedding speech, etc.), blogging, data analysis, research, and so on. These are only some of the fun things to do with chatGPT.
Answers
AI can combat climate change by improving climate modelling, optimising carbon capture, managing supply chains efficiently, assessing climate risks for finance, analysing environmental data, promoting sustainable behaviour, and accelerating innovation for climate solutions.
AI in the climate sector will create new jobs in research, data analysis, and technology development. It will automate routine tasks, increase efficiency, and transform existing roles. Workers may need new skills for collaboration with AI, and jobs related to policy, regulation, and green technology development will emerge. While AI offers opportunities, addressing workforce challenges, such as reskilling and potential displacement, is crucial.
Problems With ChatGPT
I’m glad that chatGPT raised some concerns about its potentially negative impact on human workers. Addressing those challenges will lead to an ethical use of artificial intelligence tools. Being a climate copywriter, I obviously feel a bit of pressure. However, I think that chatGPT is not and will never be able to replicate or trigger human emotions, which is the essence of copywriting. Therefore, you’ll need someone like me to make chatGPT write like a human and produce high-quality content, which is still the top ranking factor according to google.
2. Vaibhavi Joshi
Bio
Vaibhavi Joshi is the Lead Marketing Manager at Eugenie.ai and specialises in crafting innovative campaigns focused on sustainability.Â
Answers
As climate change exacerbates the severity of storms, wildfires, and droughts, AI and digital tools are valuable for predicting and mitigating these impacts. AI can help analyse vast environmental data, optimise energy usage, and manage natural resources.
AI will create new job opportunities in R&D. The automation of routine tasks will require shifting skills toward data science, machine learning, etc.
3. Thomas Basikolo
Bio
Thomas Basikolo works with the ITU coordinating and managing the AI for Good’s Machine Learning 5G activities. His interests include machine learning, deep learning and network science, and their applications in wireless networks.
Answers
Drought prediction using machine learning can help save time and costs compared to manual assessments. Machine learning is also aiding in comprehensive understanding of changes to our planet. For example, the Earth Map shows real-time information about climate challenges.
AI for Good is supporting innovative start-ups by running an open competition and awarding the best ideas to combat climate change, called the AI for Good Climate Change Innovation Factory. Start-ups with leading solutions presented their ideas at COP28.
4. Daniel Baldassare
Bio
Danile Baldassare is a climate dynamics researcher creating more reliable forecasts of changes to critical climate systems by combining atmospheric science, machine learning, and economics.Â
Answers
For climate mitigation, AI can be used to reduce emissions by improving efficiency in a multitude of systems including the grid, manufacturing, logistics, etc., and also in R&D to design products which emit less. For climate adaptation, I'd say within decades at a minimum, AI-based climate models will be able to better project climate impacts.
I am an academic, so I can really only speak to this area, but there is going to be a huge issue in the next few decades. Researchers who have deliberately chosen to remain ignorant about AI will be unable to contribute in their fields.
5. Nick Valenzia
Bio
Nick Valenzia is the co-founder of Leafr, an AI-driven platform for freelance sustainability specialists.
Answers
I don't think it's a question of solving at this point. It’s more about mitigating the worst impacts. There's a strong use case for AI to help companies fill their huge sustainability skills gap. There are parallels in medicine in reducing capacity gaps: AIs have been trained to read X Rays and accurately diagnose issues, in a fraction of the time and cost that it takes to train a radiologist. You could imagine similar use cases in the climate space, for example assessing weather risks, deforestation trends, or identifying supply chain emissions.Â
So in the micro perhaps, but in the macro I don’t think so. As a historical example, the advent of digital banking technology from the 70s reduced average sellers per branch, but the overall number of bank tellers increased, because banks cut their costs and so could open more branches. However, history also tells us there will be workers in legacy industries who will be disrupted. Think of 1800s farmers being laid off when steam-powered farm engines were introduced. This is where the concept of a just transition comes in: ensuring that workers in legacy sectors are brought along with these vast societal changes.
Conclusions
Clearly, this human-bot brainstorming session revealed that artificial intelligence has a huge potential for supercharging our efforts to stop climate change. However, we need to decarbonise AI training while retraining people who will be affected by automation over the next few years.
Comentarios