There are many predictions as to how many jobs artificial intelligence (AI) will replace. Undeniably, AI is progressively achieving human-level performances in certain tasks such as autonomous driving, food preparation, cancer detection, and game playing, achievements which have made it seem as though humans are slowly becoming obsolete. However, AI's impressive achievements only paint a one-sided picture. For example, when AI systems are left to function without human assistance, they can occasionally make mistakes or even completely fail (for instance the fatality caused by a self-driving car), making them vulnerable for adversarial situations. Because human intuition is not yet present in machine learning systems, no matter how great, intelligent, and fast our AI systems become, they will still need humans to train, operate, and supervise them.
The future convergence of real-life problems and machine capabilities will lead to the rise of what I call shared intelligence (SI). SI will occur when human and machine intelligence are combined to help overcome the limitations of existing AI systems, while enhancing human capabilities, especially in complex decision-making scenarios. For instance, the startup, Cindicator, is focused on combining the collective intelligence of human analysts with machine learning models to make better investment decisions. The company is harnessing the wisdom of crowds by aggregating analysts' insights, while developing multiple models to uncover predictive patterns from the data available.
In the future of work SI will trump the use of humans or AI alone for achieving performance thresholds never achieved before. However, the rise of SI raises some important questions: How will machines and humans effectively collaborate and communicate with each other? Which skills will be required by humans? Are our current training approaches and education models properly equipped to prepare the collaborative workforce? In order to answer these key questions, the companies must complete a paradigm shift in terms of work, workers, and the way we learn:
- Creating a new work task allocation system to achieve an optimal balance of human-machine collaboration businesses will need to deconstruct jobs and identify the tasks best performed by humans and those best handled by AI systems. Business leaders will need to create a task allocation system to define roles and responsibilities and set the rules for coordinating AI systems and human workers to accomplish a task. As more machines begin fulfilling worker tasks, businesses will require new roles, such as a man-machine teaming manager (responsible for identifying tasks, processes, systems and experiences to be upgraded by newly available technologies), as well as new approaches, required skills, interactions, and constructs. I believe augmented reality/virtual reality (AR/VR) will be a driving force for enabling workers to collaborate meaningfully with machines through a simple and intuitive interface (translating consumer behavior to business users, as well as to machines, for instance).
- Rethinking team structures. Should you integrate a human-machine team into your existing workforce strategy or develop a new workforce strategy in parallel with their existing approach? Unfortunately, there is often no one-size-fits-all answer to this question. In either case, human-machine teaming will change the way organizations manage their workforces, workflows, workspaces, and cultures. Teams of the future will be more flexible and fluid and leaders will place more emphasis on work output and value-added activities and less on the number of hours worked.
- Relearning how we learn. As a result of this new shared task framework, humans and machines will both learn new ways to improve their performance in a certain task. Machines are ready to work and learn from us, but can we say the same? In today's rapidly evolving world, we must ask ourselves the question of whether or not we are effectively teaching students/training workforces to collaborate with machines. If you are like the average company or educational institution, your answer to this question will probably be "No." In the future, every worker will need to familiarize themselves with AI systems by learning basic technical constructions and tweaking machine capabilities to exploit the full value of the system they are working with. However, our current education systems and workforce development programs have been slow to adapt to the new realities. Fortunately, we have developed a future of learning equation to help leaders and educators fundamentally redesign their learning processes and structures by responding to the changes needed to equip people for the work ahead.
The shift towards shared intelligence will take time, practice, and effort. While some companies, educational institutions, and individuals will stay the course and learn, others will just sit back and watch as their role becomes irrelevant. Shared intelligence is the future of work, so here's to human-machine collaboration!