Is AI useful for project managers?

2024-08-23

Is artificial intelligence (AI) useful for project managers?

There are some good, general uses, but

they are not specific to project management,
not as many as some people claim, and
you don't have to learn much to use them.

So, ignore the people and organizations who are trying to sell you something related to AI in project management and spend your limited time and hard-earned money on something useful.

Let me explain why.

Competency vs. AI usefulness

The following diagram shows what I think about the relationship between a user's competency in a particular domain and the usefulness of AI to that person in that domain:

We are all novices in many domains. When I have a question or problem in such a domain, I may take the short route and use one of the generative AI chatbots to get a simple answer. I know the answer is not reliable, but I'm not going to do anything serious with it. I also don't have enough knowledge in that domain to comfortably use the right resources in a reasonable amount of time.

For me, project management is different. I have the knowledge and experience that make it easier and faster for me to find and understand new reliable information and to process and analyze data. It usually takes me longer to do that with generative AI, and the result is of lower quality. That's the red part of the diagram. There are a few exceptional areas where the current AI shines. However, those are general and useful in many domains, not specific to project management. Those are in the green area of the diagram.

Practical result

The current trend in AI for project managers is all about topics in the red area of the model. Its proponents portray them as being in the green area, but anyone familiar with project management knows that's not the case. Expert project managers don't dare say much about it, for fear of being accused of being technophobes.

This diversion of people's attention is harmful. Rather than learning something unreliable (things in the red area), it's more fruitful to focus on moving from the left to the right of the model by gaining knowledge and experience.

What about the future AI?

What I've said is about the current state of AI. Things will be different in the future, but it's difficult to predict how. I don't expect a revolutionary change in the next ten years, though.

Doing everything reliably in the red area requires multiple inventions (if possible at all), and simple progress in what exists in today's AI won't be enough.

Regardless, even if AI becomes so great that it can reliably do everything in the red area of the diagram above, it's still better to wait for that to happen than to waste time playing around with today's AI toys.

Remember that you've been using AI for many years, usually without even realizing it. That's how useful AI will continue to be in the future: You won't need to learn how to use it; it will be designed to help you do what you do seamlessly.

Then, why is everyone talking about it?

Yes, many people and organizations (some famous in the community) are talking about the uses of AI in project management as if there's been a revolution or if we should expect one very soon. However, their number and fame don't guarantee that they are making sense. After all, bullshit has an incredible power.

There's also a vicious cycle: Software vendors feel pressure to add something, no matter how irrelevant, to their software just to be able to say that their software uses AI. This, in turn, leads to more people asking for such software. I hear more and more about specialized software that uses AI for things that can easily be done with algorithms.

Update 2024-08-26:

I wantched a good presentation from Jodie Burchell titled "Beyond the hype: A realistic look at large language models". You can find the video on YouTube: https://www.youtube.com/watch?v=Pv0cfsastFs

— the end —