Skip to content

Don’t ask what AI could be used for – ask how

AI offers such a wide range of possibilities for your business that it can seem impossible to choose the best ways to use it. That’s why AI projects shouldn’t focus on different options but on how to implement them, argues Sofigate’s Mikko Saari.

Have you ever spent an entire evening trying to choose a movie from the seemingly endless selection of streaming services? Have you ever simply not watched a film because of too many choices?

Don’t worry, you’re not alone. With a handful of options, choosing the best one is easy. But when faced with too many choices, people become paralysed because they can never be sure that the option at hand is the best possible one. After all, there may be an even better one out there.

Choice overload paralyses the organisation

A business often faces a similar sense of abundance when it decides to use generative AI in its operations. A company that approaches AI adoption as a traditional IT project will spend a tremendous amount of time and expertise determining what would make the most sense to use the AI for. 

However, there are too many options to make a rational decision. Thinking about where an organisation could use AI services across all of its operations can quickly produce a vast number of mixed possibilities, ranging from extensive strategic processes to small day-to-day operations and everything in between.

Difficulties can also arise in the next phase of the project. Even if a company manages to select a set of activities in which it will use AI, it can never be sure of its choices. It is easy for managers to feel that the company is turning its back on a large number of activities where AI could be even more useful. 

At the same time, the result is likely to be mixed for the functions for which staff are trained. Some employees will embrace the training and find that AI adds value to their work. Others, on the other hand, will become passive and conclude that AI is of no use to them.

The results will be significantly better if management sets strategic goals for the use of AI and then lets the staff find their own ways to make use of the new tools. By encouraging experimentation and self-learning, the company ensures that every employee uses AI in a way that is most beneficial to them. Exploratory learning is further enhanced by a culture that enables sharing and peer support for experiences and lessons learned.

A strike force pushes the agenda forward

The key point about new services is that they represent plug-and-play technology that can be implemented quickly. This ease of use means that anyone can experiment with the various possibilities of the services without limitations and can find their own preferred way of using them. The result is that two people doing the same job end up using the services in two very different ways. By sharing their experiences, each is likely to discover new and different ways of working.

This is why shackling staff with instructions and passive training in ways imposed from above simply does not work. On the other hand, using AI will not become effective based on a completely unguided experiment either. Implementation requires the support of a separate “strike force” that actively pushes the AI agenda forward. 

In practice, this means a group of enthusiastic pioneers who take the initiative to explore the various possibilities of AI, identify concrete use cases, experiment with different tools and services, actively share their experiences with the rest of the organisation and guide anyone who needs help with practical problems. 

The task force can also help shape common rules for using AI. However, it cannot take ownership of AI and its use. Instead, its role is to enable experimentation and encourage the less daring.

In other words, among the seemingly endless options for AI, it is not even worth trying to find the best ones by means of a traditional IT project or a management power grab. So, the real question in making use of AI is not what. It’s how.

Read more:

Three mistakes to avoid when implementing AI


Mikko Saari leads Sofigate’s Business Technology Studio and Operator business, which combines end-to-end IT services, business technology management services and world-leading platform solutions for mid-sized growth companies.