Understanding the distinctive nature of AI projects

- Imagine that you run marathons, you have prepared for it. You know the path and you are optimistic about what is ahead. You did something similar a number of times. Then the next day, someone asks you if you are a good runner, and then you say, "Yes, I am." Then he puts you in a 100-meter race blindfolded on a route that keeps changing. It's an entirely different race. When you manage AI projects, it's exactly the same, and I'm not talking about employing AI to improve your project management skills. Now we are talking about you managing and leading projects that will use the power of AI at the center of delivering benefits. It could be a new way for insurance companies to control risks using generative AI. It could be a new music studio that uses AI to make new music, or a tool for hiring people that uses AI to analyze and screen candidates. They move at a different pace if compared with traditional projects. But one important point I want and I need to highlight is that I am not personally a big fan of using the term traditional for projects with some level of predictability, including a clear scope and a defined and measurable set of outcomes. Traditional is sometimes related to something outdated or old or even out of fashion, and it's not. There are thousands of projects that must have a clear roadmap to succeed, and I can tell from my own experience working a lot on capital projects. Consider constructing a new airport without a defined plan. What would you do in this situation? Will you create a runway as the minimum viable product or MVP solely for testing and learning? Spend a few hundred million dollars only to see whether it works? It does not make sense, and this is why traditional project management approaches are massively helpful for thousands of projects. However, with these new AI projects, there is a lot more discovery and exploration. Predictability does not exist at all. If you are working on an AI project to optimize your business supply chain, the level of results you can achieve might only become clear as the AI system learns and evolves. It's not a predictable industry. Another aspect to be mindful is the AI business landscape. Every 20 minutes, a new relevant AI tool is released 24 hours a day, seven days a week and many people say the interval is even shorter than that. These projects evolve much faster than traditional projects and the outcomes can often be uncertain due to continuous innovation. In these initiatives, a lean structure that can decide and act quickly combined with a cultural mentality that embraces change and adapts becomes critical. In that AI project you might start developing based on a specific algorithm. One week later, yes, one week later, a new algorithm is released with a new, much more powerful and more efficient machine learning engine. What will you do? If you want to survive, you must adapt quickly. It's a whole new level of volatility.