Developing an AI and Digital-First Workforce
Tuesday, August 18, 2020
Contrary to some opinions, artificial intelligence (AI) and machine learning are not a death knell to the modern workforce. Rather, there are real opportunities for machine learning and AI to enhance leaders' abilities to make real-time data-driven decisions. As our working lives continue to change due to the impact of COVID-19, it’s become more pressing to explore the way that AI and machine learning can help us work more efficiently. If our next generation of workers learns to work effectively with advanced AI-based solutions, they will have an increased ability to work more creatively and productively than ever before.
What is machine learning and how is it connected to AI?
The key difference between modern machine learning and AI is that machine learning is not “rule-based” as older AI or expert systems are. Rule-based systems are effective when you are progressing through a series of options and each option has a set answer. Tax filing systems such as Turbotax are a great example of a rules-based system that has helped cut down time spent on filing taxes. However, most human problems are too complex to be effectively addressed by rule-based systems. In contrast, modern AI moves beyond sets of rules and can learn by itself—this type of program is only limited by our imagination to ask the right questions and provide it with the right data.
Replicating human intelligence is a challenge for AI
Teaching AI to learn in an ambiguous environment is an enormous challenge. Human innovation often comes about through a process of mistakes, adjustments, and trial and error, but it’s difficult to teach a machine good judgment. The ability to connect seemingly unconnected dots to create new knowledge is still a human gift that machines do not possess.
However, the next generation of AI-driven technology needs to be able to come to new conclusions in order to be truly useful in the working environment. With this purpose in mind, it’s helpful to think about AI and machine learning as complementing humans by discovering what humans are unable to articulate. Embracing machine learning and AI-driven applications as tools to discover what humans cannot, as opposed to replacing human judgment, opens the world up to a wide variety of uses in business settings.
Business areas where AI, machine learning, and analytics are already in use
AI and machine learning powers the world’s most sophisticated marketing operations and even modest-sized companies are collecting data and leveraging it with the help of high-powered tech. Marketing applications include recommending products or movies to customers, predicting what people may want to buy, and dynamically matching sellers to buyers.
Inventory management, usually a big cost center, has attracted a lot of machine learning applications to reduce cost, calculate risk, and improve efficiency. Human resources departments are also turning to AI-driven technology to help with hiring and employee-management.
While opportunities abound, some of these potential uses have presented ethical dilemmas that need to be addressed for machine learning to truly help us work better. Due to the “black box” nature of AI technologies, AI solutions can reinforce tacit bias or negative outcomes, reflective of the inputs they receive. Most famously, some AI-driven hiring processes have inadvertently reinforced human biases and produced unfair outcomes for job candidates.
Using AI-driven solutions to restore individual creativity
The smartest, most human-centric approach to AI and machine learning is to have the machine step in only when it deems help is needed. Over-reliance on machines is dangerous. We need human intelligence to continue to innovate, think critically, and make complex decisions. These human-driven actions are imperative for continued growth and for our own well-being.
Our biggest bottleneck to implementing AI-driven and machine learning solutions in business is understanding where these human characteristics can be complemented, but not muddied by machine intelligence.
Starting your work
The future belongs to those individuals, leaders, business units, societies, and countries that optimize the man-machine duality. AI won’t replace managers, but managers who use AI will replace those who don’t. Discovering how machine and human intelligence can complement each other in specific use cases is key, because although some areas will benefit from machine intelligence, not all will. As you begin to consider how AI-driven solutions can be used more broadly, you’ll need to focus on identifying which business processes or applications are suitable for an automated environment and which are not.
Associate Dean of Faculty and Research
Curtis L. Carlson Chair in Information Management
Professor, Information Decision Sciences
Alok Gupta is the Associate Dean of Faculty and Research and Curtis L. Carlson School-wide Chair in Information Management at the Carlson School of Management, University of Minnesota. He received his Ph.D. in Management Science and Information from the University of Texas, Austin and teaches courses in the areas of computer networking, electronic commerce, decision support, IT infrastructure, and computer programming.