Artificial intelligence, events and sustainability. I guess it’s time to talk about it.
In recent months I’ve been approached by groups wanting feedback on their AI-driven sustainable event tools.
Imagine my thought bubble at being asked for 30 minutes of my time to “pick my brain” about the development of tools designed to replace what I’ve done in a human-to-human way for over 25 years.
The use of AI in the sustainable event space is inevitable, and may even be advantageous. My students are already using AI tools to reduce learning friction. They generate lists of sustainable site selection and catering practices faster than I can teach the topics. Not only that, but they are thinking with it and using it to apply knowledge. They create prompts that help them critically analyse the quality of things like ecolabels and carbon offsets. It even helps them understand and spot greenwashing.
AI is helping my students quickly assimilate what to do and what not to do when planning a sustainable event.
Still, I am unsettled by some aspects of AI’s growing role in the sustainable event field. And not just for the obvious reason of the environmental footprint AI has, or the risk of becoming personally obsolete. I wonder about things like:
Can AI really help plan a sustainable event? To some degree, I think yes. The landscape of sustainable events is technically complex, and getting more complicated by the week. And while we have more human helpers in this field than ever before, AI can be used to efficiently scale-up knowledge-sharing to meet growing needs. Well-designed prompts drawing on good, robust content can help organisers design a pathway that makes sense for their unique event, quickly prepare communications and aid in measurement.
What can language models miss, and our prompts fail to uncover, when we use AI in service of sustainable events? I recently tested an event AI tool to generate sustainable travel tips. Using the default prompt provided by the tool, I was quickly presented with ideas related to public transit, active transport, electric vehicles and carbon offsets as ways for the event to “significantly reduce its carbon footprint” from travel.
Helpful for a strictly local event. But my operating context is international events, and it took two prompts of my own to find tips related to the most significant source of emissions in this case: air travel. And none of the tips provided included exploring virtual options to attend, considering train travel, or using carbon footprint tools in flight booking applications to evaluate the impact of different itineraries.
In another experiment I asked for advice on choosing quality carbon offsets. The tool provided some pretty good criteria, explaining the need for permanence, additionality and verification. However, it failed to flag the shortcomings of offsets, and neglected to list the consideration of Indigenous consent, which is a critical point of responsibility for stakeholder groups I work with.
Although there was a tiny sense of relief I hadn’t been completely replaced yet, it was hardly a failure.
What the exercise did highlight is that AI for event sustainability operates in a rapidly evolving space where event and user contexts and stakeholder concerns can be quite unique, sensitive and fluid. And that avoiding blind spots relies on cultivating the user’s ability to ask skilled questions while also thinking critically about the results.
Who controls the foundation model for sustainable event AI and how does ownership influence information? While the user’s role and skill are critical, there is also a need to be aware of who is funding and what sources of information are used to populate AI models. This is particularly important for sustainability, where active tension between stakeholders and demand for credibility and transparency drives continuous improvement.
To put it plainly, a sustainable event AI has the potential to provide different recommendations depending on if it is paid for and/or trained by an environmental non-profit, virtual event provider or destination.
So, accepting that AI will inevitably be funded by interests that may have an agenda, it is important to consider how inclusive foundational models are, and if outputs draw on the knowledge of leading edge, objective specialists.
Is AI technology more burden than benefit? For AI to credibly support sustainability in events (and any field), the emissions reductions brought about by its use must outweigh the environmental burden resulting from its operation. It must also cultivate and improve talent, rather than reduce it. Which brings us to the next and perhaps most critical question:
Can AI help clarify what we are really being called on to do to be sustainable in events? Can AI cultivate the personal skills and abilities that are necessary to solve challenges? And nurture an ongoing commitment to really improve?
In asking I’m assuming that sustainability in events fundamentally requires us to contribute to human flourishing within planetary boundaries, through our event goals, formats, logistics and outcomes. (If you are unclear what I mean by that, this WEF video may be helpful.)
And that requires qualities like courage, creativity, humility, empathy and persistence.
I acknowledge that might not be everyone else’s assumption. After all, the sustainable event AI tools I’ve looked at so far don’t seem designed to deal with difficult questions that interrogate the fundamental need for the event, or it’s scale and location when it is deemed worthwhile to host one. They’re largely geared toward helping “green up” business as usual event models with self-generated goals and checklists based on what we’re willing to see and prepared to think about.
To use an example, AI can provide suggestions for building a green booth. It can prepare a checklist to select a sustainable venue. It can even recommend creative CSR projects for incentive trips. And help track progress to reduce waste and emissions. But it doesn’t (currently) prompt the question: “Is the event product fundamentally sustainable?”
And AI doesn’t help navigate the anxiety and ambivalence we feel when such questions surface, from within ourselves and others. That’s up to us.
My purpose in sharing these thoughts is not to advocate against using AI in event sustainability practice. More that we think carefully about how we use it, and be aware that it has the potential to add impact, and distract and delay us from asking important questions related to our future, and the role events play in achieving sustainability goals.
Let’s not let AI fool us into being content with what is convenient to make events more sustainable. Let’s be open to the possibility that if AI is making things easier it may be a sign we’re engaged in status quo thinking, and avoiding questions that cause productive discomfort.
And when we do tap AI, let’s use the efficiencies created to dedicate ourselves to the highest sustainability potential of what events can do: creating experiences and communities that are brave enough to ask difficult questions and inspiring human-to-human collaboration to solve tough challenges to secure a livable future for all.
There’s no AI substitute for that. At least, not yet.
Eventcellany 