In this article, we will discuss what I believe is one of the most significant issues facing the future of project management. Today we’re looking at the impact of AI and machine learning on the future.
Big management.
Let me start by asking three questions.
- Why was the date the 5th of December 2017 a momentous day for AI and machine learning?
- What is exascale computing?
- What does any of this have to do with project management?
If you’re a project manager and don’t know the answers to those three questions, I suggest you read further because your career might depend on knowing them.
- Why was the date the 5th of December 2017 a momentous day for AI and machine learning?
So why is the 5th of December 2017 an effective date for those who take even a cursory interest in the development of AI and machine learning, or ML? The 5th of December 2017 was a pretty special day; on that day, one computer beat another computer at the Top Chess Engine Championship. Now I know what you’re thinking. You’re saying, so what? Computers have been winning at chess ever since Deep Blue beat the chess grandmaster Garry Kasparov on the 10th of February 1996. Well, first of all, let me say I’m impressed. Not many people know that, but why the 5th of December 2017 was so important?
Is it the Rocky Balboa, Apollo Creed Matcher of the computer chess world? So let’s get ready to run more in this corner.
We have stockfish 8, a computer coded with hundreds of thousands of open-source coding hours, a program designed specifically for one purpose: to play chess.
A computer with computing power to run through a Monte Carlo Tree search of 70 million potential moves per second and the reigning world champion having come first or so. Taken in nine out of the last 14 seasons.
And in another corner, we have A0, a computer that has never been programmed to play chess, a computer that’s been given no access to previous information or end game tables, and it’s only been programmed with the game’s rules. A computer that is, in fact, self-taught with the capacity to only search through 80,000 moves per see, and A0 was given 13 hours to teach itself the game. During those 13 hours, A0 was only allowed to play against itself and run an internal analysis to review its performance. On the tournament day, A0 played 100 time-controlled games against Stockfish 8. The result was a winner by Knockout A0 with 28 wins, zero losses, and 72 draws. So A0, an AI-based computer, had used machine learning to beat a traditional program comprehensively. So how did it do this?
Well, in layman’s terms, it used intuition instead of logic. Instead of crashing through millions of options, it learned from its experience using a plan to check and act, the process for continuous improvement. It made value judgments. Based on its own experience and decided which of the options presented the best chance of success. In other words, A0 started to think like a human in terms of machine learning. This was a massive step forward. So hold on to that thought cause we’ll come back to it. Let’s move on to the next question. What is exascale computing?
What is exascale computing?
This is the fastest computing that the human race has ever developed.
An exascale computing system can do a quintillion floating-point calculations per second. That’s 10 to the 18th power. So surely, that type of power will be available to most users in the distant future. And even now, the US government is spending $1.8 billion on developing three exascale supercomputers, which they anticipate will be operating by 2020. And the US is in a race against China which already has three working prototypes, and Japan and India, which have committed billions of dollars to develop exascale computers within the next three years. So you tell me, is it something that’s really in the distant future.
So what does all this mean for project managers?
Since the inception of the Project Management profession, we’ve been shaped by the Iron Triangle. When you try to explain what a project manager does, you will sooner or later revert to using the Iron triangle of time, cost, and quality to explain your role. This simple idea has served us well for decades. But will it serve as well in the future? Let’s look at how AI and machine learning can impact the core roles of planning, monitoring, and controlling projects, and let’s start with time.
Based on what I’ve told you about AI and machine learning progress, do you think it’s within the realms of possibility that a computer could interpret drawings and specifications? Schedule out a program monitor that schedules in real-time and then make adjustments to the forecast and reprogram to achieve an optimal critical path. If so, how long do you think it will be before we have computers that can do this? Is it five years away? Maybe only three years away?
Try last year. I recently watched a demonstration of a first Gen PMA. I wear a computer using only the bin model. Schedule the entire construction program now. When you couple this with the rise of Lidar-enabled drones and site-based bots with the ability to traverse every aspect of the site to capture and monitor progress, then the idea is that the computer can complete most of the time elements of project management is already becoming a reality.
Now I can hear you say yes, but what about using that information to assess contractual issues like extensions of Time claims? Surely AI can’t do that.
He’s an interesting fact.
AI is already being used to predict legal cases at the European Court of Human Rights with a 79% accuracy. AI is doing legal research and undertaking legal due diligence in mergers and acquisition deals. It can identify errors in legal documents and even test legal arguments against recent case law and existing legal precedent. Do you still think the PM’s ability to review and assess EOT claims is unsafe? OK, so that’s time gone from the triangle. What about cost?
Well, unfortunately for us project managers, that’s the easy one. Computers linked to the Internet easily attribute a cost to materials, transport, and installations when coupled with their self-generated program. It’s feasible. To believe in an AI-based program or know exactly when to order what product using machine learning functionality, it’s not unreasonable to see that same computer able to search the Internet for the best available price, send the order to the optimal supplier, and ultimately reduce the cost of construction.
So if you’re counting on your skills as a cost estimator to keep you in the game, all I can say is I’m sorry. That leaves us with quality.
How about smart glasses that building inspectors can wear to assist them with building inspections, checking for compliance with plans, and even referencing the quality standards based on a minimal database. Still, it won’t be long before they don’t need a person to walk around the job site?
Why not just set these smartglasses up like a CCTV? Or better yet. Build them straight into the building itself so they can be used to monitor the long-term health of the building. Unfortunately, that’s quality gone. So there you have it. AI and ML are undertaking all the activities associated with the Iron Triangle within the foreseeable future. AI and ML are taking it all, except for a few traits like Ethics and Moral Leadership in Project Management. Ethics and Morals in Project Management is altogether a different beast for AI and ML to command.
So what does that mean for us who’ve invested so much the time in perfecting the skills and competencies used to achieve the Iron Triangle definition of success, on time, under budget into the required quality. So, I think we’re standing on the cusp of a significant shift in the Project Management profession. Don’t get me wrong; I don’t think we will achieve a Skynet-type singularity event anytime soon. But we need to start a serious discussion about how we will redefine, retool and retrain ourselves very soon. Or we risk becoming blacksmiths in the age of the automobile.
So what should we do about it? Let me know?
