The Future of Productivity: AI and Machine Learning
Grow Your Business, Not Your Inbox
The productivity and project management market is booming, and it’s continuing to evolve in new and exciting ways. I wanted to know what the future of artificial intelligence in project management would look like, so I reached out to founders, productivity experts and futurists who work in this space every day to ask what their predictions are for the next five and 10 years. Their answers were enlightening.
We'll use AI for decision support, not decision making.
David Allen, the inventor of Getting Things Done, believes, ”Systems will get better at presenting the relevant data to optimize our experience in every situation -- at the right place, at the right time. We need to think of productivity systems as supporting systems for our decision process.”
Related: How to Prepare Employees to Work With AI
So, we won’t yet be using AI to eliminate our decisions and automate them, but to enhance the ability to make a decision in any situation. I also think this is the next likely step for AI. Most people wouldn’t trust a computer to make decisions for them, but they do look for information to help make those decisions.
Allen goes on to say, “Neither these new presentation forms nor trends like A.I. will make a decision for you. That won’t work. I see that A.I. can support your decisions but we still will use our heads to make decisions.”
It’s not robots . . .
When most people think of AI, they think of robots doing manual labor to get things done for us, whether it’s Rosie from The Jetsons or a robot on the assembly line at Ford. The issue I have with this is that this view is too limiting. Sure, robots will become intelligent, but so will everything else. The team behind the online video series In a Nutshell released a great explainer video about this.
Right now, an app that will be able to recognize your motivation level and give you tasks that fit that level is not out of the question. Bots that can answer simple service questions and learn from the responses are already around, working with some success and some big failures.
Mark Mader, CEO at Smartsheet, thinks that thinking of AI as roving robots is missing the point, saying, “Looking further out, there’s no doubt that automation -- don’t think robots, think removing mundane and unproductive work steps from your day -- will increase. Machine learning will be able to predict what workers are trying to do and make their work easier. How? By automatically gathering the information they need to complete a task, populating forms and sharing them with the appropriate people.”
I see a combination of both. Specifically in project management, I see a future where machines will be able to predict a change using real-time data and make changes accordingly. It’s the combination of bots and machine learning that holds the key: Think of an assembly line system pushing out barbecue equipment. The system will be able to predict that demand will increase due to an upcoming holiday and automatically tell the bots on the line to increase production. I’m not sure if there will be a human decision between them, but I think as we become more comfortable with the machines’ decisions, we’ll give them more control of the process.
It’s my decision to make.
AI is far from being human, let alone superhuman. As I pointed out, we’ve released machine learning bots with some pretty terrible results. Machines are not yet good at understanding context or sarcasm, so when we let them learn, they usually miss the mark.
Using machines to help with a decision, however, seems like the only way forward for the moment. As a form of decision support, productivity expert Carl Pullein thinks that “machine learning and artificial intelligence [will move] towards creating productivity tools that can schedule your meetings and tasks for you and to be able to know what needs to be done based on your context, where you are and what needs to be done.”
Machine learning enabled tools like Grammarly are already on the market, but as these decision-making aides become more well-known, they are moving into more complex areas. Think of it as your Facebook timeline algorithm or your spam filter, but for your to-do list.
Just like education, access is the key to machine learning.
To get a sense of where we are now, think of systems that are on and learning all of the time. Your computer browser talks to Google and Facebook and any number of companies, where what you’re doing, clicking, buying and beyond, is stored. Now with the recent trend of IoT, these things are moving away from our computers and cellphones and becoming part of our everyday lives. And gaining access to data along the way. Your smart fridge might be able to tell the local supermarket how much water you drink per week. If everyone in the local area has a smart fridge, that same supermarket would be able to make a better decision about how much water to keep stocked.
Productivity psychologist Melissa Gratias sees this system working for us in our workplaces too. “Most apps and programs require the user to purposefully interface with the tool in order to use it. We will see more smart homes, smart cars and voice-activated entry points that allow the tool to be always available to the user, no matter where she is. She won’t have to stop what she’s doing to, for example, add something to her task list.”
Related: Will a Robot Take My Job?
So, not only will the supermarket know what to stock, but your fridge will know what to add to your shopping list for the next week. Automatically, through learned behavior.
All the experts I spoke with agreed that decision support is likely the only way forward in the short term. Machines are being taught the decision preferences of humans, because they can’t discern context on their own. So, we need to educate machines on what it’s like to think like a human. Companies like Alphabet are already working on it, with projects like DeepMind at the forefront of AI and machine learning technologies. Others, like Elon Musk’s OpenAI, are working to make sure that humanity’s fears of a malevolent AI will never be realized. As we learn to trust these systems, adoption will quickly follow. And since they're so universal, they will surely touch all industries.