Forget Taking Over the World: Will AI Destroy Us With Carbon Emissions?

You’re probably bored of the discussion over whether AI will take over the world and destroy us. If you’re like me, you’ll think the conversation is pointless. The world is already being destroyed by the less sexy topic of our own carbon emissions.

But amid climate crisis fatigue, environmentalists are waking up to the fact that the environmental impact of generative AI is huge. In a few years, an individual AI system is likely to consume more energy than a single country.

Is AI the most environmentally destructive industry we have? No. But if I were an XR campaigner, I would target AI ferociously, hoping the world might actually pay attention.

So how do we as marketing and tech professionals who use AI in our day-to-day respond to this? This blog post will cover:

• The environmental impact of AI 
• How to use gen-AI responsibly
• Achieving brand sustainability and authenticity

What’s the environmental impact of Generative AI?

The training process for large AI models involves running computations on powerful hardware for vast lengths of time. This uses significant amounts of electricity and contributes to carbon emissions. You also need a lot of water to cool the processors down – in the case of ChatGPT’s Iowa data centre, 6% of the district’s.

The production and disposal of hardware components like GPUs and data centres depletes recourses and generates electronic waste.

As AI forerunners start training more and more complex systems – and as more companies get involved – the environmental toll will be unsustainable. Sam Altman acknowledged this himself at the World Economic Forum in January 2024: ‘We won’t get there without a breakthrough.’ 

@bloombergopinion Did tou know that #AI uses an insane amount of #energy and #water ? @Kyla Scanlon #nuclear #microsoft #chatgpt #climatecrisis #waterwaste #samaltman #oklo #terrapower #billgates ♬ original sound – Bloomberg Opinion

Is using ChatGPT bad for the environment?

A typical Gen-AI request will use about 5 times as much energy as a search on Google or another search engine. Naturally, image and video requests use more.

Training artificial intelligence appears to be the bigger issue. The energy it takes to train a large AI model is equivalent to the energy of around 200 million interactions with it. But we have to factor the cost of creating the beast into the impact of using it. 

So, can you claim that your organisation ‘puts sustainability first’ and still use ChatGPT or Gemini? Keep reading.

The positive potential of Generative AI

Many believe that AI could be the key to solving the climate crisis. This is largely down to its ability to read large amounts of data in mere seconds and predict significant patterns. These results can help us produce power more efficiently and prevent climate-related problems before they occur.

Like with any new technology, you’ll hear phrases like ‘the possibilities are endless’.

But do the benefits outweigh the cost?

I don’t know. I’m a content writer in a Shoreditch-based digital marketing agency.

What’s clear to all of us is that, on a planet on the brink of collapse, the AI industry needs to create its systems more sustainably. Like what Hugging Face is trying to do with its large language model BLOOM.

OpenAI’s approach

Sam Altman, CEO of OpenAI, believes AI is heading towards an ‘energy crisis’. He also thinks the AI revolution is going to get us on a path to near-zero.

The proposed solution to the energy-eater in the room? Nuclear fusion power plants. Altman has invested a lot in Helion Energy, one of a few companies attempting to create an environmentally friendly power source. All as yet unable to produce more energy than they use in the process of creating it.

How to use AI responsibly

Ultimately, it’s up to governments and tech top dogs to regulate our use of AI. But if you’re a business making claims about sustainability on your site, it’s important to take some personal responsibility.

Here’s how to do it.

1. Educate yourself and your team. Stay up to date with the latest research into AI’s environmental impact.
2. Push for change. Reach out to your MP – or even to those with influence in the industry – to request tighter control of AI’s environmental impact.
3. Exercise a human-first approach. Use AI as a tool when it will be most helpful to you but prioritise human-created work where you have the capacity to do so.
4. Share your stance publicly. Explain to clients that you treat generative AI as a costly resource. Share this article on LinkedIn!

Using AI as a sustainable business

These days, there’s a cost-benefit analysis in almost every decision we make. Just as you might decide a long-haul flight is worth the footprint if the trip has a good cause, you’ll find many uses of AI where the benefit outweighs the cost. 

At Digivate, we utilise AI within our human-first approach to all areas of digital marketing and development. If you’re looking for support marketing your sustainable business, get in touch today.

Article by

Tess O'Bamber

Tess entered the world of marketing as a content manager, focusing on social and blog content. She now has over four years of experience writing on-brand copy for websites, emails, whitepapers and more. Tess has a Master's in Creative Writing from the University of East Anglia – she takes a creative approach to copywriting, whatever that looks like for each unique client. You'll still find her working on her first novel at the weekends, between DIY sessions and long walks.