Project management and AI: current issues

Could you introduce yourself and tell us about your role within the Groupement romand de l'Informatique (GRI)?

So the Groupement romand de l'Informatique, GRI is the leading professional association for ICT professionals in French-speaking Switzerland. We have around 500 members, and I often say that these are companies, not individuals.

What's interesting is the diversity: historically, there were a lot of software and hardware players... but today there are also companies that use IT, so practically every sector.

The GRI also plays an important institutional role: it is recognised as an organisation of the world of work (OrTra) at federal level. In practical terms, this means that we have the legitimacy to represent the sector, particularly in matters of vocational training - apprenticeships, federal certificates, etc.

(smiles) And it's an organisation that's been around since 1971... so it's seen the arrival of computers, and everything that's followed.

For my part, I represent the GRI on certain bodies, chairing a committee and representing the industry's interests in public discussions, including political ones.

My background is fairly cross-disciplinary. For a long time I worked at the frontier between business, technology and public issues. I started as an international civil servant at the International Trade Centre, an agency linked to the UN and the WTO, working on international trade and development issues.

Then, since 2017, I've been working in the private sector, with consultancy and training assignments on concrete projects, often with a strong technological component.

What always comes back is this role of “facilitator”. Or interface, if you like. I translate between worlds that don't always speak the same language: technical profiles, business decision-makers, public players... The challenge is to get them to work together.

In organisations that are often overwhelmed by day-to-day activity, how can you successfully manage projects based on their real impact?

The classic trap is to be very busy... without necessarily moving in the right direction.
What really changes the game is a very simple question: what needs to change in the end? If this answer is not clear from the outset, the project will quickly go off in all directions.

Then there's the discipline. With so much to do, there's a tendency to add tasks, to respond to everything... but you have to keep coming back to the essentials: does it really contribute to the objective?
(slight laugh) Easier said than done, of course.
Successful projects don't necessarily require more resources. Above all, there's the ability to stay focused on the impact... and to eliminate anything that doesn't add value.

And at what point do you realise that a project is drifting off course?

Often, in the beginning, everything goes well. There's even a kind of euphoria. Everyone is motivated.
The role of the project manager is to maintain this momentum... while remaining clear-headed. And then there's the reality that we often underestimate: the balance of power within the organisation.
There is sometimes a tendency to over-value the individual... but in practice, a project is also about teams that can have tensions, marketing versus sales, for example.
The project manager has to navigate this. Understand the real dynamics, not just the theory. Today, with experience, I'd say we need to go back to something more concrete, less “theoretical”. Many models have shown their limitations.

Working together has become more complex with teleworking and the multiplication of contacts. How do you get everyone on the same page?

For me, there is one non-negotiable point: important issues must be dealt with in real time.
Feedback, criticism, a sensitive question... if it's only done in writing, by email or chat, it can quickly create tension.
It really does. It can even make situations worse.
You need to call, do a video, or ideally see each other.

Have you ever been in a situation where the written word was a problem?

Yes, more and more, in fact.
Before, we had more of a reflex to pick up the phone. Nowadays, I see a lot of short messages piling up... and sometimes a kind of distance creeping in.

(pause) I'll be blunt: I've even noticed a certain “cowardice” in some of the written exchanges.
So we have to compensate by recreating dialogue in real time. Even 10 minutes can be enough to clarify a situation.
And often, the problem isn't the distance. It's that everyone has their own version of the project in their head. So you have to realign regularly: rephrase, check that everyone is talking about the same thing. It sounds basic... but that's where it all comes down.

Structures evolve rapidly. How can roles and responsibilities be clarified over time?

When things aren't clear, it's immediately obvious: decisions drag on, nobody really makes up their minds...
Rather than complicating things with highly theoretical organisation charts, it's better to start from the ground up: on a key moment in the project, who decides? Often, the answer exists... but it's not shared.
I'm still fairly convinced of the usefulness of hierarchies, without falling into rigid models. Completely flat structures, in my opinion, create other problems, but project management goes far beyond the organisation chart.
You have to mobilise people who are not under your direct responsibility. And that's where we come in: leadership and communication.
Over time, what counts is not a perfect model, but a comprehensible and stable framework.

With the development of agentic AI, how can the division of roles between humans and technologies evolve?

There are a lot of fantasies surrounding AI... in both directions.
AI is highly capable of analysing, structuring and producing information rapidly. It can even perform certain tasks autonomously. But it doesn't take responsibility. And it doesn't have the full context.

The real question remains: who decides in the end?

Today, AI can already take on some of the preparatory work, suggest options and save time. But as soon as there are strategic, political or relational issues at stake, the decision remains a human one.

A recent use of AI that stood out for you?

Some companies are starting to offer services such as auditing and consultancy, entirely via AI platforms, at costs divided by 500, sometimes 1000.

And it's clear that the model is changing radically. Jobs that used to take weeks to complete, with junior teams... are becoming automated. It's only a matter of time before this becomes widespread. It's bound to have a major impact on certain professions, especially those at the start of their careers.

At the same time, it democratises access to services that were previously reserved for large companies. And that's a major change. By way of comparison, today's development is probably faster than that of the Internet in its early days.

So yes, the profiles that will come out on top will be those that know how to use AI as a lever... while retaining a real capacity for decision-making. Because, ultimately, the more technology advances... the more human responsibility becomes central.