It seems like every other day we read something new about the rise of artificial intelligence (AI) and the ultimate demise of post-industrial work as we know it. It used to be that stories about factory workers being replaced by robots dominated the news. Now, in 2017, its driverless vehicles, clerk-less retail outlets and digital professionals such as lawyer-bots and medical-bots that make the headlines.
Yet, as project managers, we tend to ignore the media noise regarding the coming robot apocalypse with a sense of, “Considering what I do for a living, I could never be replaced by AI, right?”
This article explores the possibility that project management as a profession may not be as protected from displacement as we’d like to believe. And even if we don’t believe that replacement is possible, the impact of AI on our field can’t be ignored.
The Rise of the Intellectual Bot
Before we can comprehend the future impact of bots, we need to understand just what a bot is. We hear about bots all the time. But how could something programmed be intellectual — in other words, think and act like a professional?
The term “bot” was once synonymous with “robot.” Today, instead of referring to a physical entity with an AI brain, capable of running around on all four mechanical limbs, the term bot refers to a digital activity manager.
A chatbot is the perfect example of this new beast, existing only in binary characters and appearing on our device screens — instead of standing behind the reception desk of, say, the Henn na Hotel in Gamagori.
Bots vs. robots: Bots only exist on computer screens, while robots appear as receptionists or even dinosaurs
A chatbot engages the user via a chat window or audio sensor (think Microsoft’s Cortana, Amazon’s Alexa or Apple’s Siri), asking questions and capturing data on the transaction, then feeding that knowledge to another bot, a software application or a human being to perform a relevant action. If you have ever gotten product support via a chat window on a website, you have engaged with a chatbot; or perhaps you are already using a bot to order Amazon products online (via Alexa and her physical presence, the Echo device).
You can see one example of a professional-leaning bot on DoNotPay, a legal chatbot that blatantly attempts to cut human lawyers out of routine legal proceedings.
Example of a bot
The DoNotPay chatbot attempts to cut the lawyer out of legal proceedings
Developed by then-teenager Joshua Browder, this project employs big data from IBM’s Watson technology to deliver legal options deduced from a user’s natural request, such as, “I was just arrested, so what should I do now?” The idea is that that common folk can become their own lawyers, by being given all the relevant filing instructions, forms, templates and advice to win their cases. A BBC article published last year claimed that the bot had helped 160,000 people challenge parking tickets since its launch. Earlier this month Browder updated his bot to allow the many millions of Americans who have had their personal information hacked in the Equifax data breach join in on a class-action suit; all a user needs to do is go to the homepage and click on “Automatically sue Equifax for $15,000.” The bot takes the user in hand from there; no human interaction needed.
The Future of Professional Bots
“Thinking” bots are showing great promise with IBM Watson technologies leading the way. For example, take the bot called Ana, soon to be helping folks navigate insurance-related quagmires using natural language analysis and Watson’s big data reserve.
This insurance-bot will help folks navigate this complex industry without waiting on hold or pressing “9” to speak to a customer rep.
And this tech is sure to improve as AI continues to infiltrate any profession that includes routine, mundane or repetitive tasks. As a PM, can you think of any of those?
The PM-bot, Theoretically Explained
While the “PM-bot” has yet to be built (at least none have shown up yet on chatbots.org), it’s only a matter of time before developers begin integrating bots into software that PMs depend on for their everyday activities.
For example, Microsoft recently formed Microsoft Research AI, a dedicated unit focusing exclusively on how to make the company’s software smarter. The unit is already applying AI to sift through emails to collect enough data to deduce intentions, such as when and where to hold meetings and to perform other office tasks, such as making a personal commitment to perform a particular task:
Microsoft Cortana’s “Commitments” feature can pull the promises you’ve made out of an e-mail and keep you to it!
So, when thinking of all the mundane and repetitive tasks that we do as project managers (scheduling work, evaluating costs, determining risks, balancing resources, etc.), is it possible that these familiar PM activities could be accomplished by a bot?
I think so. Consider the nature of the modern-day project manager when starting a new project: asking the big questions and needing the right answers. I’m sure Watson would be up to the challenge or — more pertinent to Microsoft Project users — whatever Watson-like equivalent comes out of the Microsoft Research AI unit. Any such bot should be able to: 1) deduce the type and scope of the project that you are planning; 2) determine what resources are needed; and 3) calculate if there were budget to fund the project — just based on your asking!
Then, once it knew what needed to be done, the PM-bot could: 1) produce an array of possible schedules that meet the requirements; 2) include any risk mitigation ideas based on past and like projects; and 3) inform and interact with a rightly-formed team to get the work done, all the while tracking project burn-down.
Sound like science fiction? Well, we all know what most sci-fi turns into after a period of human ingenuity: reality.
The architecture for a PM-bot can be visualized below; and it’s only a matter of time before a PM-bot is built by an inspired software development team (with a shout-out to the Project developers):
A possible PM-bot architecture, including Microsoft Project technologies
We have to wonder — once this bot exists — what the duties and responsibilities of the human project manager will be — especially in a world where machine learning, AI infrastructure and collaborative bots could enable anybody to handle project management.
At a minimum, I imagine, these bots will “augment” our PM reality, allowing us to focus on the softer skills we humans provide, leaving the bot to do the more mundane aspects of our work.
The End of the Human PM, or Just a PM Evolution?
It’s clear from these few examples of current bot technology (and a rudimentary architecture that almost exists today), that the duties and the responsibilities of the project manager as we know it will change. Does that mean we should all run out and find a more secure position — perhaps as a bot developer?
Well, as with driverless cars and drone-delivered groceries, the jury is still out. We’re not seeing human doctors replaced by Dr. Watson yet, so perhaps we’re all safe for the moment. And that’s because AI and bots can’t yet replace what humans do best: make mistakes gracefully and then use our out-of-the-box thinking and often illogical creativity to make better decisions — thus discovering even better ways to manage projects.
Image Credit: NASA
Learn how an MPUG Membership helps individuals and teams become better project manage professionals and Microsoft Project users through highly cost effective online training.
Join MPUG to attend live training webinars, access 500+ hours of on-demand sessions, receive certificates of completion and earn the Project Management Institute (PMI)® Professional Development Units (PDUs) that you need to become a Project Management Professional (PMP)
Watch an MPUG training webinar for free and improve your Microsoft Project skills in less than 1 hour.
[ezcol_1half]
Project Management Resources
- Online Training
- Start Your Project Management Career
- 3 Tips for Successful Project Managers
- Scrappy Project Management, Part 1
- Scrappy Project Management, Part 2
- Laying the Foundation for Leading a Project Management Office
- What is Microsoft Project?
[/ezcol_1half][ezcol_1half_end]
FREE MPUG Resources
- Microsoft Project Tutorial
- Tips from a Microsoft Project MVP
- Webinar: Microsoft Project Do’s and Dont’s
- The Project Communication Plan
- Creating an Agile Schedule
- 15 Tips for New Users
- MPUG Newsletter
- Want access to more? Join MPUG Today!
[/ezcol_1half_end]
Raphael Santos
Very nice article Jigs!
And, in fact, there is somebody already working on bots for Project Management. Check it out:
Christy
I think PMs should certainly keep an eye on AI and not assume they are safely isolated from the coming “robot apocalypse”. In the types of projects I do, I can (and have) automate a fair 80% of the job with the simplistic tools available to me, which leaves me ample time to do the “value add” such as cultivating excellent stakeholder relationships, supporting project team members, and preparing for multiple possible scenarios as the work moves forward. If I can do this with MS Project, Outlook, Excel and some VBA…I have to imagine the world is wide open if some enterprising individuals really put their minds to it. (Such as the PMOtto.ai folk)
David L Leech
interesting article Jigs.
Microsoft have also added Bots to the Microsoft Teams product.
– The T-Bot is a bit like the old clipper, an interactive help tool, and is available already.
– The Who-Bot is way more interesting, it will search for references across your connected infrastructure (OneDrive, Intranet, Azure, email, etc) for whatever you are looking for.
– And they provide an api for you to build your own bots.
https://www.onmsft.com/news/microsoft-teams-introduces-t-bot-and-who-bot
Eric Uyttewaal
Jigs,
Great article! I had never even considered that bots could take the place of PMs.
I still think they cannot do it, because there is too much human interaction needed with a variety of stakeholders, not to mention the motivating and influencing we do. Also, having reviewed 1000s of schedules and still finding incomplete and incorrect network logic in 90% of them, has convinced me that bots will not be able to produce schedules in an automated way as you suggest. After all, if humans cannot do it, we will never be able to teach it to robots. Perhaps, I should get into that … 😉
However, I do see how these robots could enhance a project manager’s productivity significantly, so we will need less of us … eventually …
Thanks again!
Eric Uyttewaal