This article is part of the Forum Network series on Digitalisation, feeding into upcoming discussions at OECD Forum 2019, and the 20th anniversary of the OECD Forum, contemplating its past and imagining its future.
The OECD Forum has reached its 20th year; this is a good moment to reflect on where our emerging digital, artificial intelligence-driven society will be in 20 years when the Forum celebrates its 40th.
The conventional wisdom is that artificial intelligence (AI) is spreading like wildfire through the economy. But there is little wisdom, conventional or otherwise, on the exact skill sets necessary to educate and train the emerging AI-driven workforce. There is however an emerging consensus that we need soft skills as well as hard.
More and more global companies like Deloitte and LinkedIn point to a serious soft-skill gap, providing reams of data on, and analysis of their benefits in the workplace, as well as the costs of their absence and the difficulty finding them. McKinsey claims companies leave USD 800 billion to USD 1 trillion on the table annually for lack of employees with the necessary interactive skills. According to the Wall Street Journal about 90% of executives surveyed report difficulties finding these necessary skills. And the US Labor Department announced that “Soft Skills Pay the Bills”, and hence are necessary for all.
Hard skills and soft – not yet in sync
If there is agreement we need multiple skills in the digital workforce, the two big rubrics – soft skills and hard skills – are not yet in sync. Exactly how much of each do we need? How to integrate hard and soft? Does everyone need the same skills? What’s the mix of basic and specialised skills? And can you even teach soft skills, or people skills or interactive skills? Certainly the answers to these questions will shape AI’s diffusion around the globe.
The shape of the future is now in the hands of those educating and training tomorrow’s workforce. Not only is there a need for those who will use AI on a daily basis in factories and laboratories, there is also a huge demand for those who design those systems. The educators training the system designers need a model, a “vision” of what the workforce will need tomorrow. They require a vision of how AI should work in multiple social settings and for multiple purposes, while meeting the needs of multiple populations.
Identifying core soft skills
At the University of Southern California, we believe we have found some answers to these questions. Following seven years of extensive, multi-method research our team has identified a portfolio of soft skills widely sought by employers around the world – competencies to guide choices about AI and digitalisation more generally. The five essential attributes are Adaptability, Cultural Competence, Empathy, Intellectual Curiosity and 360-degree thinking.
We call this model Third Space Thinking (TST), a model and mindset quite distinct from two other leading paradigms: those of business schools and engineering schools. Borrowing from those fields, though, we introduce greater rigor, consistency and conceptual clarity than other definitions of soft skills. Our goal is to “put hard edges on soft skills” by making TST rigorous and better integrated, carefully designed to be effective in practice. The model is a communication-based “soft technology” to solve problems in the digital economy.
An example from AI’s future
What kinds of soft skill challenges and opportunities does AI face today? The consulting company Accenture has concluded that in addition to eliminating jobs, AI in the workplace is also creating new ones, some of which did not exist in the recent past. They are emerging, but the dots haven’t yet been fully connected.
Accenture identifies trainers, explainers and sustainers. These “uniquely human jobs” are differentiated by the work they do in the spaces between machines and people. Trainers train machine algorithms to better understand human communication. Explainers “design smart behaviors” based on “particular business contexts…and individual, professional and cultural factors”. Sustainers include automation ethicists and others who work directly with people to better inform them how AI works and can be sustained.
Let’s try a couple of thought experiments. Imagine three paths to the future of AI. In one, diffusion accelerates quickly. In a second, a graph of AI diffusion shows it spiking, then it drops. In a third, we see slow and unsteady expansion.
One of the main determinants of these possible paths will be the pace at which we mesh the different kinds of skills required to meet AI’s very complex challenges and recognise and seize its emergent opportunities.
The rate of change of AI expansion – whether fast, medium or slow – will be shaped partly by the level of professional preparation of the people working on it, and the values and priorities they bring to their work. The “soft technology” model of Third Space Thinking can help advance the various kinds of “hard technology” on which AI draws. The hard skills and soft, like the hard technologies and the soft, should be deployed so they are complementary and advance common goals.
Put in terms of the Accenture example, the trainers, explainers and sustainers need TST attributes. The designers need lots of intellectual curiosity to seek unconventional solutions. They need insight into the broadly defined “cultures” of different communities and become empathetic to their priorities. To understand the full range of change drivers they need to use 360-degree thinking.
Let’s indulge in one more thought experiment. Consider a future where AI is designed and deployed by people with low levels of empathy and cultural competence, who don’t score well on assessments of their intellectual curiosity. Imagine that future. Now imagine an alternative future where AI system designers possess those competencies. A future where this soft technology is widely available to those who design, develop and diffuse new technologies. A future where the average person, the typical consumer and citizen also possesses Third Space competencies. That is a future we could look forward to with anticipation, rather than dread. We need all the designers, trainers, explainers and maintainers to see themselves as – and so become – the reliable and ethical stewards of our futures.
- What do you think the balance between hard and soft skill training should be?
- How could we integrate hard and soft skill training? Or do you know of any examples of this?
- Can you even teach soft skills, or people skills or interactive skills? And does everyone need the same skills?
Continue the conversation and help us co-create the agenda
||New Jobs & Occupations||OECD Forum 2019|
Find out more about OECD Forum 2019: World in EMotion