When AI leads to organisational redesign
Your company has experimented with generative AI. You know how fast and capable the frontier models are, how readily they produce convincing answers to almost anything you ask. The potential is real. You can feel it. The people in your company are working in new ways and faster than before.
And yet your business goals remain unaffected. So where does the value from all this AI use and development investment actually go?
It goes to individuals completing single parts of value creation, and to pioneers exploring and developing point solutions. They report new capabilities, decisions grounded in a broader base of information, and faster iteration on ideas, but they also have more and more outputs to test and review, which often end up on other people's task lists as well.
So individuals have become more independent. They carry the operational risk of AI themselves and treat their AI literacy as a personal asset. In an uncertain economy and a global rush toward an AI-fuelled wave of efficiency, that is exactly where their incentives point.
It has come at a cost. These same people have faced the frustration, detachment and cognitive fatigue of testing, studying and trying to translate vast and ambiguous potential into something the business can actually leverage. The shift shows up as the endless testing of AI features in both new and old tools, all while the principles for governing AI are still being built and rolled out. The result is a decoupling between the organisation and the people using AI within it: individual practices have sprawled and advanced well beyond anything the business has formally adopted.
Companies are trying to squeeze profitability calculations out of a malleable and shifting technology in order to gain a sense of control over the direction they have chosen and the decisions they have made. To this end they build metrics at every level, from individual performance and changes in how time is spent all the way to team-level and business-level benefits. The typical way to design this measurement is to scale it against AI efficiency targets. Efficiency is of course a baseline requirement, because when everyone in the AI race has the same tools, it is ultimately customers who cash in the gains. At the same time, we have not seen the wave of layoffs that AI efficiency was supposed to produce. If anything the direction is the opposite, with building AI-based competitive advantage on everyone's lips. Everything is moving faster, and leaders are trying to climb onto the back of the new AI workhorse.
In our research with OP Pohjola and Accenture, 65% of leaders at large and mid-sized Finnish companies believe AI will significantly change the competitive logic of their industry within the next five years, and 43% believe that over the long term AI will create entirely new business models. So many already sense that efficiency alone will not be enough, and that what is needed is standardisable, scalable advantage along with bold experimentation inside the organisation.
As Marcel Proust wrote: "The real voyage of discovery consists not in seeking new landscapes, but in seeing things with new eyes." That is exactly what the next stage of AI maturity demands: less chasing of the newest models, more pausing to consider how your organisation creates value. The question becomes how to embed AI into the structure of the organisation so that the larger context becomes clear to both the users and the models themselves. The next wave is already converging on this: purpose-built lightweight agents, context-building integrations, organisational redesign tied directly to AI. The architecture of organisation-level AI is starting to take shape. The same malleability that makes large language models so capable becomes a durable advantage only when it is fitted as a purposeful and dynamic part of your organisation's operations.
The boldest companies are setting out to solve customer problems with AI at the consumer interface. Last autumn we helped Beely build exactly this kind of context, where data, human-centered understanding and a multi-agent architecture built for solving tightly defined problems came together in an inspiring way. Chatbots tend to have a mixed reputation in the consumer world, but we see that here too: combining the consumer context with AI's operating principles already at the design stage is the missing piece, the thing that turns a generic chatbot into an element that genuinely keeps smoothing the path to purchase. The result works not as an anthropomorphic machine, but as a deliberate conversational tool with clearly articulated capabilities.
At Noren we help make sense of this direction, bringing the social dimension into the machine. Using co-design methods, we bring together the organisation's internal and external thinking into strategies and solutions that are tailored to real needs.
Further reading: Unlocking AI value in Finnish organisations
INTRIGUED AND WANT TO LEARN MORE? GET IN TOUCH!