As a ‘living-systems’ organisation, we cultivate and hold space for collective intelligence, exploring the issues of our time by bringing together diverse and independent perspectives in discussion.
How fitting then, that we recently turned our ‘collective intelligence’ to the subject of artificial intelligence (AI). Here we share the journey of our discussion; from initial mild panic at the possible implications for our role through to excitement at the potential it has to accelerate the transition to a regenerative economy.
Areas where AI is already making a contribution to sustainability
AI is already showing exciting potential in the food sector. Wasteless, for example, is using AI-powered dynamic pricing to help supermarkets and online food retailers recapture the full value of their perishable products and reduce food waste. In a similar vein, AI is now beginning to play a role in food design; fellow B Corp NotCo is using AI to prototype different recipes derived from an algorithm programmed to produce novel flavour combinations, resulting in meat alternatives that taste as close to the real thing as possible without the carbon footprint. The benefits of AI have also begun to surface in the agricultural industry where the improved ability of farmers to analyse soil data, weather patterns and crop health information is leading to reduced chemical usage, more efficient water management and reduced impacts on nearby ecosystems.
AI has also started to play a role in biodiversity monitoring and the data-informed management of forest areas under threat of illegal logging. For example, the charity Rainforest Connection has developed an innovative acoustic monitoring technology that uses old mobile phones and AI to protect vulnerable ecosystems by detecting threats in real-time. The solar-powered acoustic devices used for threat detection, called RFCx Guardians, transmit audio to the cloud. AI then analyses and classifies the data collected to detect signals like chainsaws, vehicles and gunshots, that could alert ground partners of potential threats.
AI also has huge potential to be applied as a supply-chain mapping and management tool where it has already started to contribute to the improved traceability of raw materials and labour. The rapid gains we are seeing in data collection and analysis, particularly when AI is integrated with blockchain technology, provided companies, auditors and even consumers improved visibility and authenticity of what is happening upstream in the value chain.
Tech startup Unibloom will apply AI to the great conundrum of data management in corporate sustainability. Vineet Ahuja, Unibloom’s CTO, told us it’s all about optimisation: “our software will apply AI and machine learning to business and planning decisions in order to help organisations prioritise the right initiatives and grow in a way that is beneficial for people, planet and profit. It helps businesses build credible [de-carbonisation] plans while also achieving cost savings. We harness AI to help find the pathways that allow for sustainable business growth.”
In all, AI’s potential role in spearheading the shift to more sustainable practices and consumption patterns sounds promising.
Some questions for us to consider
It is key that we seek to take stock and mitigate the negative social and environmental implications that are already beginning to surface.
Many AI models are trained and deployed on highly energy-intense servers and the sector’s ravenous carbon consumption has been subject to increasing public scrutiny in recent times. Less known, however, is the enormous water footprint associated with cooling warehouse-sized data-centers and servers. One recent study has estimated that a conversation of between 20 and 50 questions with an AI chatbot will consume around 500ml of water.
Another less-known negative consequence of AI is that of ‘AI hallucinations’ and the subsequent spread of misinformation. The term describes erroneous or unrealistic output by an AI system, particularly in the context of images, text and copy generation. Put simply, the creative nature of generative bots means they have a tendency to produce statements that sound factual, when they actually aren’t. This is particularly worrying when we think about the proliferation of generative and automatic scheduling tools in the marketing industry where AI bots, often without human oversight, may go on to produce unsubstantiated environmental claims and misleading information about a brand’s practices or product credentials. Apply this to the context of greenwashing or ESG reporting and there is a material risk to progress that needs to be addressed. Perhaps the industry currently devoted to creating such content will be redirected to creating the instructions for AI to produce it.
We must also acknowledge that the economic and social benefits of AI remain geographically concentrated in the Global North. The WEF has predicted that, while all regions of the global economy stand to benefit from AI (with North America and China seeing the largest GDP gains), countries in the Global South will experience more moderate increases due to the much lower rates of adoption of AI technologies. Without concerted efforts to cultivate an equitable and enabling operating environment, Disparities in AI readiness will feed into global inequality. In addition, steps need to be taken to ensure the data used to train generative AI models and algorithms is representative and unbiased, otherwise its use will serve to perpetuate rather than dismantle existing systemic bias.
AI as a tool, not the solution
As our ecosystem wrapped up its discussion on the role of AI in sustainability, we generated a few conclusions:
- AI is not going away. It has much potential to collect and use data better and more efficiently to inform strategy, make policy decisions and innovate for an uncertain future. We all need to befriend and adapt to it, but with eyes open to both its risks and opportunities
- Technology is by no means a silver bullet for the social and ecological challenges that we currently face; first and foremost change must be driven by shifts in our collective consciousness, the regulatory context and ultimately an acceptance that our extractive economic model cannot be sustained on a finite planet
- AI may struggle to localise and bespoke answers in a way consistent with living systems principles.
And finally…we asked Chat GPT what the role of AI in sustainability consulting would be. Obviously, it told us its role would be “crucial” but it was also kind enough to acknowledge that “human expertise and considerations” would still be needed. How kind. We won’t hang up our hats just yet.
