CHAPTER 8
The AI Advantage: Near-Term Workforce Opportunities and Challenges

By Barbara C. Matthews1

1Founder and CEO, BCMstrategy, Inc.

Backdrop

Periods of profound technological transformation traditionally trigger at least as much angst as excitement. Recent efforts to make sense of how advanced technology will disrupt and reformulate our societies draw from the nineteenth century by referencing The Second Machine Age1 and the "Fourth Industrial Revolution".2 This is not just hype. Rapidly accelerating process automation from artificial intelligence (AI) systems will generate economic dislocations even as they improve materially how people think and work.

Policymakers are reacting strategically. At their June 2019 summit, the Group of Twenty (G20) embraced these changes in their Osaka Declaration3 and issued non-binding AI Principles4 to guide policy development. G20 leaders thus seek to accelerate the rate of AI adoption within economies while committing to "empower people with the skills for AI and support workers for a fair transition".

Companies and policymakers must act now to help people benefit from the enhanced cognition delivered by AI systems.

Enhanced Cognition: The Good News

"Enhanced cognition" refers to the improvements in analytical functions from the machine-human interface. The premise is simple: computers perform repetitive tasks (information acquisition, organization and visualization), delivering outputs that provide a more advanced entry point for humans to conduct higher-order analysis.

Early technology delivered enhanced cognition through printouts, followed by digital spreadsheets with automatic calculations and automated data visualizations (charts, graphs, etc.). AI systems extend the frontier by automating basic reasoning and generating predictive analytics.

Two parallel technological advances are now revolutionizing how knowledge workers operate. Firstly, innovations in hardware and processing mechanisms (neural networks, cloud computing), expand the availability and capacity of automated analysis. Secondly, natural language processing (NLP) and smart, connected devices are exponentially expanding the amount of data available for use in AI systems.

For example, NLP makes it possible to convert unstructured verbal data into structured data. This data can be used by AI systems to identify correlations (pattern recognition) across concepts faster and better than humans. This pushes the perimeter of process automation into analytical tasks currently performed by knowledge professionals.

The promise of enhanced cognition is real for those ready to embrace it. Knowledge professionals are today able to perform more interesting and creative analytical functions. The result is a substantial increase in the velocity of insight formation by humans, as well as operational efficiencies due to decreased time spent acquiring information.

Considerable economic gains will accrue particularly to early adopters. Increasingly, AI-as-a-service is delivering advanced insight formation to mid-sized companies seeking advanced insights to help drive internal productivity and efficiency enhancements, as well as superior customer service and even public policy risk analysis.

Macro-Trend Analysis of Workforce Challenges

But reality is not tidy at the margins of this virtual, AI-enhanced world. Rapid job creation within the technology sector expands the skills mismatch, redistributing jobs (and joblessness) in the economy. This places pressure on (i) older professionals and (ii) younger unskilled individuals who are unable - or unwilling - to adapt.

This dynamic has been a fixture of the modern economy for decades, as process automation reached critical mass in advanced economies. Structural shifts hit unskilled labour first. Secretaries and typists, automobile mechanics, and assembly line workers were next. The AI revolution today places comparable pressure on knowledge professionals. Bookkeepers, stock analysts, and economists are currently on the front line of process automation. It is unclear where the disruption will stop.

Cross-border connectivity and globalization will drive dramatic shifts in trading relationships and job distribution, further fuelling concerns regarding economic dislocations. Fear of the unknown often fuels counterproductive, emotional backlash, as the current trade policy landscape illustrates.

Pragmatic Suggestions for a Way Forward

With governments seeking to accelerate the transition to a digital, AI-powered economy, companies and policymakers should focus on policy priorities that empower individuals to keep pace with - and better understand - the opportunities presented by AI technology.

1. Investments in Lifelong Learning

In order to retain talent and innovate internally, companies must expand investments in educational benefits for all employees.

The last significant technological shift in the workplace for knowledge workers occurred in the 1990s, when information processing became ubiquitous. Employers at the time financed training to teach employees how to use word processors, spreadsheets and email.

Some executives grumbled about being asked to perform secretarial tasks. But they adapted. Secretaries were asked to perform more interesting work, like database management and executive assistant functions. A similar investment in on-the-job training is needed today.

Rather than feed fears of being replaced by AI systems, companies should train employees on how AI systems function. These employees - who know the data and the business far better than any AI system - will be poised to deliver enhanced value and new insights when working with (rather than against) AI.

Such on-the-job training need not be disruptive or expensive. As the International Finance Corporation (an arm of the World Bank) highlights here,5 online courses like Udemy are partnering with companies to provide relevant offerings to workers on their own schedule as well as more formal in-person training.

The key is to empower professionals to experiment with how AI can enhance existing skill sets and job functions. Meaningful training should be interdisciplinary. Subject matter experts should become familiar with coding and neural networks. Professionals with expertise in coding should take higher-level courses to accelerate their ability to operate AI systems.

Low-skilled adults require different - but no less important - training. For example, the European Union in 2018 launched an "Upskilling Pathways"6 initiative to fund education for low-skilled adults to acquire minimum levels of literacy, numeracy and digital skills. As Germany's apprenticeship system illustrates, multiple avenues for skills enhancement exist beyond the educational system.

Retraining is good for business. Companies that invest today in upskilling their workforce will receive early mover informational advantages and operational efficiencies. Just as importantly, they will increase retention rates from employees eager to explore new horizons within a corporate environment that invests in their skills.

2. Coding as a Second Language (CSL)

Equipping people for success starts in school. Much has been written about the importance of teaching creativity and problem-solving. However, sales and marketing staff, operations management and strategy analysts must be proficient in the language of coding and the structure of AI systems in order to deploy creativity and problem-solving skills effectively.

From elementary school and up, instruction and experiential learning involving coding languages and computer processes should exist across the curriculum. This would equip the next generation to understand the strengths and limits of AI-powered processes, while providing the foundation for the next round of innovation.

Policymakers remain focused on strategic evolution. For example, in 2018 the United States Congress passed bipartisan legislation (the "Strengthening Career and Technical Education for the 21st Century Act") providing federal funding and other support for state and local government programmes delivering career development and technical education.

Private entities are also taking leadership roles. Organizations like Girls Who Code as well as innovative partnerships (see this partnership7 between the Panasonic Foundation and the Hispanic Heritage Foundation in the United States) are delivering technology skills training to underserved communities.

Conclusion

AI solutions will revolutionize business analytics in the same way that word processors and spreadsheets revolutionized business processes. The responsibility for meeting the challenges and opportunities presented by this technological shift is shared by individuals, companies and governments alike. How we choose to shoulder this responsibility will determine the rewards we...