The world as we know it today is developing at a rapid pace, where technological advancements appear like wildflowers, transforming our technological landscape. One of the most actual developments has been that of Artificial Intelligence (AI), with tools like OpenAI’s chatbot ‘ChatGTP’ gaining a bigger user base by the day, including consultants not shying away from using it. With AI being able to perform lightning-fast idea formulation and data analytics on big data, it has become interesting for consultancies to explore developing their own AI tool.

However, as with every new and hot topic that seems to conquer our world and revolutionize it with its benefits, also AI comes with certain complications. Firstly, a complication that is often not spoken about is the energy usage of AI. The development and maintaining of AI systems comes not only with financial costs but also with environmental costs. Secondly, the aspect of human factors should not be overlooked. Especially with the work of consultants, in which thoroughly considered solutions are formed through a deep and detailed understanding of the problem and situation at hand.

So, an important question arises when thinking of the future of consultancies in light of AI implementation. Should consultancies explore the possibility of using AI to take over their workload and take the environmental effects for granted? To answer this question, I interviewed Ton Metselaar, experienced management consultant at a successful business and technology consulting firm. The conclusion is that we cannot let this rapid development run on its own without looking at the consequences. Every innovation should be assessed on its environmental impact and ways to limit the energy usage should be explored. Furthermore, the human factor in consultancy work is crucial due to their expertise and relationship with the client and this should stay at the core of the value they bring to clients.

Exploring the benefits AI can offer

The speed and completeness of the answers AI bots currently provide is impressive, to say the least. The convenience of having a little AI helper when doing your work can help to initiate ideas and save time when doing routine tasks. But the future lies further, where consultancies will try to reap the benefits of using AI to analyse huge chunks of data within seconds, where normal data analysis would take significantly longer. These ways of saving time come with saving costs as well of course, making it interesting to explore the possibilities.

Unravelling the environmental threats of AI

Yet, where people picture a world where AI can perform all their tasks and we do not have to do any work ourselves anymore, it is not without drawbacks. Like other technological advancements, AI models have a certain carbon footprint. The humongous amount of data that is stored by data center servers requires a lot of energy and water usage to run the servers, equipment, and the cooling systems. Ton Metselaar explains that it’s like calculating the transactions for Bitcoin, which is known to consume large amounts of energy. AI models produce CO2 emissions in three ways. The initial training of the model is the first and most energy consuming way, with researchers calculating the CO2 emission at 626,00 tons of CO2 which can be compared to the CO2 emissions of 119,000 cars (Strubell et al., 2019). Furthermore, the carbon footprint of keeping the generative models running and letting people interact with it are slightly below these emissions. The last way is through the updating and tailoring of the model to a consultancies’ specific information, which uses the least amount energy.

Analysing the importance of the human factor

Next to the fact that AI has certain drawbacks, it also contains flaws. As Ton Metselaar accurately puts it: “AI can make a lot of suggestions and have ideas on how to make things simpler but as a model it can never be completely without mistakes”. AI might interpret certain factors wrong or miss out on contextually important details which can lead to wrong outcomes and bad solutions. It is therefore not a reliable source. In addition, the absence of human empathy and understanding is also lacking with AI since they are unable to genuinely understand the client concerns. Clients always seek an emotional and human understanding that AI cannot offer. This can make clients feel like they are treated as data points rather than individuals with unique requirements and specific concerns to be addressed. AI cannot establish trust or build reputation with clients through meaningful conversations, active listening, and the human touch that comes from genuine interactions, which are the cornerstones of client relationships.

Proposing a sustainable way of AI usage

But what should our future look like then?

Ton Metselaar continues to explain that for a good consultant, it is essential to be an expert in their field. Therefore, it is vital that they can carefully listen to what a client needs and what specific situation they are in. Subsequently, this information should be mindfully analysed to come to the optimal solution. Ultimately, this needs to be properly communicated to the client, with which the relationship and the maintaining of that relationship is indispensable. All these aspects show that AI cannot replace the human factors consultants bring to the table. Furthermore, with our world being threatened by drastic environmental issues, “We should not only save the planet, but also save ourselves. The climate will likely kill us before it will kill the planet”, as Ton Metselaar alarmed in the interview. When looking in further detail at the solutions we should not aim for every consultancy to build their own specialised AI model but rather build on existing ones to limit energy usage. Moreover, it is needed to incorporate AI usage in the monitoring of a company’s emissions to be aware of its magnitude. To really seek for responsible improvement, environmental-friendly AI developments should be sought instead of creating completely personalized ones.

Reference list

Kumar, A., & Davenport, T. (2023, July 20). How to Make Generative AI Greener. Harvard Business Review. https://hbr.org/2023/07/how-to-make-generative-ai-greener?autocomplete=true

Strubell, E., Ganesh, A., & McCallum, A. (2019). Energy and Policy Considerations for Modern Deep Learning Research. Proceedings of the AAAI Conference on Artificial Intelligence, 34(09), 13693–13696. https://doi.org/10.1609/aaai.v34i09.7123

Acknowledgement Statement

This blog is part of the student writing competition in Management Consulting Master Program at the School of Business and Economics