You can’t win the AI race by running the old track
Speed is still a competitive advantage when renewing a business with AI. But even that is of no use if we keep running on the old track and focus on making current work more efficient, says Sofigate’s Aleksi Halttunen.
Dear business leader, feeling out of breath? No wonder. At this time of year, there is plenty of that going around. There may be many reasons for it, from the approaching next quarter to changes in EU regulation, but one is fairly common: trying to keep up with the AI conversation.
AI is now talked about as a race. The winner is the one who adopts the latest model and technology first and gets the most out of it. The pace of development is almost impossible to keep up with, and it challenges even us experts.
Even so, I would argue that the most dangerous illusion in the AI race is not that we are not running fast enough. The greater danger is believing that we can keep running on the old track without rethinking work, processes and leadership.
Three out of four companies are automating the old
For our latest playbook at Sofigate, we examined how companies are moving from AI experimentation towards the age of digital workers. Our study combines observations from practical client work with the views of more than 60 companies operating across the Nordics and the Benelux.
The findings show that 73 per cent of companies are still using AI primarily to automate existing tasks, rather than to rethink work itself. It also reveals a common error in judgement: AI is still being viewed mainly through the lens of cost savings and process optimisation. The focus remains on how current work can be done faster, more cheaply and with fewer resources.
That view is far too narrow, because automating old work is not transformation. If we improve one or two stages of a workflow, the bottleneck simply moves elsewhere.
Don’t optimise. Maximise.
AI’s real potential begins to come into view when we shift our gaze from optimisation to maximisation: what would a thousand or ten thousand digital workers make possible for us? Could we serve new customer segments, make better decisions, shorten lead times or build services that would not have been realistic with our old capacity?
At this point, it is worth clarifying that by a digital worker we do not mean a single bot, but a role embedded in a workflow, with an owner, boundaries and a measurable task. When business-critical processes are redesigned to work alongside them, much more than the workflow itself changes. Job descriptions, skills, decision-making, responsibilities and links to other processes all evolve as well. That kind of change is unlikely to be easy for any company. It is laborious and, at times, painful.
Even so, we have to adopt a kill your darlings mindset: we must dare to challenge familiar and previously successful business models, and let go of old assumptions about work, capacity and leadership. In the software sector, for example, we have already seen how competitors born into the AI era gain an edge because they can build their operations from the ground up on a new logic. You do not catch a new agile player while carrying the weight of the old.
A new division of labour: human + digital worker
What is slowing companies down above all at the moment is a lack of ambition in how they think about AI. In fewer than 20 per cent of the companies in our data, people are actively discussing a future in which there are more digital workers than human employees. Even that scenario can be seen in two ways: have digital workers replaced people, or has the company grown by that amount?
Based on my own experience, I do not believe AI creates value by replacing people. In customer service, this is already an everyday reality in many places: customers often still want to deal with a person, but the digital worker operates in the background. It can retrieve information from different systems, prepare a solution and trigger the necessary processes across organisational boundaries. When a human then evaluates the situation and makes the decision, the customer can receive genuine added value: a fast solution based on better data and smoother flows of information across the organisation.
And what about an engineer designing an industrial solution or a construction project for a client? If a digital worker brings together the client’s needs, lessons from previous projects, procurement constraints and the downstream effects of later stages already at the design phase, better design decisions can be made earlier. That shows up later in procurement, lead time, cost and, ultimately, in the value experienced by the customer.
In the knowledge work of the future, one cannot exist without the other. There is no human doing knowledge work without AI. Nor is there a digital worker that is not guided by a human: someone to provide context, interpret weak signals, supply material and make decisions. Even the best technology is only as good as the data it is fed. Do you really want your data to be produced entirely by AI?
The most important decision a leader will make
The most important question for any executive team right now is not whether to adopt AI, but where in the business’s core processes it should be used to build new competitive advantage. In other words: where do we direct the new capability and capacity that digital workers bring, and who carries the business responsibility for it?
The governance and responsible use of AI are not matters only for IT, compliance and legal. If there are hundreds or thousands of digital workers, they can no longer be managed as isolated agents or as experiments confined to individual teams.
AI that genuinely renews the business and does so responsibly requires clearly defined ownership, clear boundaries, supervision and logging, as well as shared metrics. In other words, it calls for a centralised governance model with these elements built in from the outset.
That governance model is worth building now, while processes are being redrawn anyway.
Download the playbook: There is no more work without AI
Would you like to read more about how digital workers can be led, scaled and put to work in business renewal?
About the author
Aleksi Halttunen leads Sofigate’s Disruptive AI business. He helps companies renew their business, value streams and leadership for the AI era. Aleksi is particularly interested in how digital workers are changing the logic of work and how organisations can use them to build new value instead of merely optimising the old.
Why most ServiceNow transformations fail before they start
“When hard work is no longer enough” – how six young talents are building AI tools for youth employment
