Guest blog (AtkinsRéalis): how can AI bring happiness to our work?

AtkinsRéalis’ Technical Director, Jim Green, uses the Japanese concept of Ikigai to explore how AI can bring happiness to our work.

Happiness, Ikigai, and AI

We are happiest in our tasks when we feel that our output is valuable, when we can be creative or challenged with complex puzzles, and when we work well with the people around us. There are numerous models for how happiness can be categorised, one of which is Ikigai, a Japanese concept in which happiness is seen as a result of four categories: doing what you love, doing what you are good at, doing what the world needs, and doing what you can be paid for. Theoretically your career should offer you all four of these elements – AI is one of the routes to make this more true.

A much-hyped fear-of-the-moment around AI in the media is the impact on professionals, the fear that an individual’s value is being removed in favour of content created through text-based requests in Generative AI.

However, if we flip that view to consider the many other forms of artificial intelligence currently in use, opportunities quickly become apparent for an individual to become more creative rather than less. A professional with AI at their disposal has more time for greater specialist work and fewer drudge tasks (increasing the feeling of ‘doing what you love’), application of their most detailed skillsets (increasing the feeling of ‘doing what you are good at’), an opportunity to focus on optimising their results (‘doing what the world needs’), and critically, the ability to do all of this rapidly and therefore cost-effectively (‘doing what you can be paid for’). AI is achieving this right now across a range of professions.

An immediate example applicable to a wide range of professions, but picked out here in ecology, is the ability to automatically identify points of interest on photo, video or audio recordings. Ecologists rely heavily on this footage to evidence the presence of species. Watching the footage, however, frequently results in a lengthy task, as you scrutinize a mesmeric b-roll of film triggered by leaves rustling in the wind, without any fauna in sight. This is both expensive, tedious, and delays action to design a positive impact for the subjects.

To avoid this, at AtkinsRéalis we have introduced AI in the form of a deep learning YOLO v5 neural network. The power of GPU computation means the solution can process many days of video footage within a few hours, successfully identifying all instances of wildlife. This automation yields a ten-fold increase in productivity for ecologists who are freed from performing the process manually. For an individual ecologist, this now means their time can be redirected to expert intervention design, utilising all of their expertise and creative skill rather than simply staring at footage. Their role becomes more interesting, more challenging and, as their time is not ‘lost’ on the initial phases of finding the species, they can join the rest of the project team more rapidly – offering their information at an earlier stage and therefore getting the satisfaction of having their views taken into account in as streamlined a way as possible.

This last point, of a project or programme team working well together, is another reason to look to AI for a source of future individual happiness. A complex infrastructure project in design, construction, or operation phases requires a small army of people, each working on a particular specialist area. Bringing these individuals together, blending their timelines, interdependencies

and even vernacular, is a constant task throughout a project lifecycle. Delays in one specialism can throw everything off and produce plenty of stress all around. Here, AI can give us a stressbusting power-up to our ability to estimate risk, by providing a way for planners on large complex construction projects to proactively prioritise activities based on project risk, and to gain a better understanding of overall project schedule health.

At AtkinsRéalis, we put this into practice by embedding a predictive analytics solution into our PMO suite. Machine learning is used to predict the likelihood of task delay and impact on project critical paths, and models have inbuilt retraining programmed to respond to performance drifts. And these tools are put directly into the hands of our programme professionals; Large Language Models (LLMs) enable planners to use natural language to query project data and automatically return intuitive results in a format that works best for them. That’s an instance of Ikigai improvement both for the planners and for the rest of the project; they can all be confident that their specialist tasks (which they love and are good at) will be blended in the most optimal way to achieve the best result for the overall programme (to ensure they’re producing a result the world needs, while also staying in budget). And there’s far less stress, thanks to the enhanced ability to foresee and manage risk.

So, how can you ensure that AI is implemented to produce happiness in every case? Firstly, by considering the strategic goals of your AI before implementing, and the specific results you want to obtain, with a view to the end users’ needs. In both examples discussed above, our AI professionals worked closely with the relevant teams to ensure the AI concept worked for them and would scale effectively. Ikigai might be a way to open that discussion, but perhaps we take a particularly mirrored conclusion here, and think of simply applying human intelligence as a pre-requisite to artificial intelligence. Either way, happy trails!

techUK - Putting AI into Action

techUK’s Putting AI into Action campaign serves as a one stop shop for showcasing the opportunities and benefits of AI adoption across sectors and markets.

During this campaign, techUK will run a regular drumbeat of activity, including events, reports, and insights, to demonstrate some of the most significant opportunities for AI adoption in 2024, as well as working with key stakeholders to identify and address current barriers to adoption.

Visit our AI Adoption Hub to learn more, or find our latest activity below.