Episode transcript The transcript is generated automatically by Podscribe.
Hello, everyone, welcome to the Five Minutes Podcast!
Today and next week, I want to do a super tiny series of two episodes about artificial intelligence in projects in what I want to do. I want to discuss with you the aspects, the use and benefits of AI in the project environment. Next week, I want to talk about the challenges. And I just want to anticipate the challenges are not very easy to overcome.
But let's start by the beginning. What is AI? Do you know one thing that concerns me it's every single person that uses a computer, sometimes thinks that he or she is doing AI. And it's not, it's not because you were using a computer that you were using artificial intelligence. Artificial Intelligence is the branch of computer science that aims to create machines that are able to do tasks that are typically only available to humans. So we are talking about tasks that require the kind of intelligence that so far, only humans are able to learn because we have natural intelligence. This is a cognitive aspect that we developed over 1000s and 1000s and 1000s of years, AI tries to simulate that, try to understand this. Many people think that AI machine learning, it's all the same thing. So AI, it's the broad discipline, machine learning is one of the possible ways for you to reduce human behaviour. So machine learning, it's like a subset of artificial intelligence. And deep learning is a subset of machine learning that basically use the concept of Neural Networks. For example, in 2015, I wrote an article at that time, nobody was talking about artificial intelligence in projects. And I use a mechanism of neural networks to help you to access and evaluate your portfolio of projects, in order to decide and forecast the potential results of your projects. And this is a tiny and specific use of AI.
Basically, AI we can see pretty much everywhere, and it's growing dramatically, for example, self-driven cars, a car that drives by itself. How does this work? It's because there is an artificial intelligence process in place, that try to mimic in some ways the human behave the driver's behaviour, in terms of pressing brakes, turning right turning to the left, stopping turning signs in these is what makes you create a self-driven car. Another example that is pretty much everywhere has at home. These tools like Siri, Alexa, these kinds of virtual assistants, where you talk to people, and they answer in answer your questions and comments on that. The other one is about chatbots. For example, many times you have a chatbot that you are able to talk to and to communicate in a simple way. For example, you send an SMS to a bot and the bot replies to you and you can do activities that are typically handled by humans, like Help Desk, for example, another use, call centers. Many times when you call a large call center and you start talking with the other side, you don't know if it's recorded or if it's a human on the other side. Many times you only realize that you are talking to a recording or to work computer just later or maybe when you make a question that is a little bit tricky to be answered. So this is AI. But now how AI can be applied in project management because this podcast is not about AI in general.
But it's aI applied to project management. Because AI is basically a fantastic tool to monitor patterns using large scale data. It's excellent to support admin tasks, for example, updates on schedule, setting reminders, bots, this kind of bots, I just told, in order to reduce the time many project managers or team members, no matter your delivery approach, they always complain about oh, I'm spending a lot of time in admin tasks. I'm spending a lot of time approving the budget, I'm spending a lot of time doing I would say budgeting and I'm not having too much time to really, really concentrate on what should I do as a project manager or as a scrum master. So these kinds of AI helps you to do that. For example, I did with partners, we created a chatbot. That was a fantastic experience for me called PMOtto. And the PMOtto was a chatbot. That was your project management assistant. Basically, what do you start a WhatsApp or a Facebook message? And you start saying, I want to start a project. And then you start answering questions and talking to Otto. That was something like you're our automatic planner. And at the end, you have your plan ready. Do you say I want to create a project? Or what is your activity in your project? Do you have a deadline or not? And do you have a budget? What is your aim? What is the value you're planning to deliver? Are you working alone on this? Or do you work with a team, and based on each of these Otto started learning about building your schedule, and at the end, you have your schedule populated or your JIRA populate or your Trello boards populated. This is a perfect example of the use of artificial intelligence to support you with administrative tasks.
Another use is the budget and schedule forecast. Many times you want to say look, which are the patterns that drive costs up. So basically, the computer can learn and say every time I combine this kind of job with this kind of location with this kind of resource, I increase the budget. Every time I combine these plays with that complex with that risk, I reduce the budget. And using that you can forecast very nicely if your project has a chance to be over budget or under budget or delivering on time, even when you are doing a business analysis to see the viability of the use of software, for example, you develop a feature and you apply this feature in two or three cities. And based on the results of these two or three cities, you have very strong advice that you should move to another city or that you should just help that project because it will not deliver the benefit you want.
Finally, you can use these 14 fittings, for example, you can see what is the best combination of professional workers and resources like machinery and equipment
that will give me more predictability. So you may say if I combine on that sprint, this type of team with that type of work, my success rate goes from 50% to 85%. If I combine this team with that work, for example, the performance will go down 30%. This is the kind of activity that is just human-related. Of course, this seems scary, because now we are talking about, will be hired or interviewed by AI software, and not the human. I don't know the technology today. But maybe in the future, these may happen. And I don't know if they are already using that. But these are clear use of artificial intelligence.
And last but not least, identification of risks evaluation of risks. Instead of you doing the group and teamwork to identify the risks, what you can do, you can give hints to a chatbot, for example, or a mechanism like Siri. And then after that, you will receive a pre-populated risk register with the most typical risks based on previous experiences. And based on learning. Of course, all of this sounds like magic, right? It's just a perfect word. But it's not. Because the challenges are extremely big. It's very easy, quite easy for me to say this to you about AI. But it's not easy at all for you to put this into practice. And this is what I want to discuss next week with you. It's about the challenges, but I want just to anticipate one thing, it's how you can have reliable and validated data that will be the basis of your learning. And this is exactly what we will discuss next week.
I hope you enjoyed this podcast and see you next week with another five minutes podcast.