AI adoption and the future of work
Artificial intelligence (AI) technologies are already disrupting working patterns across the economic spectrum. Forward-looking businesses, educators, and workers simply can’t afford to ignore the impacts and potentials of the new industrial revolution.
What’s at Issue
Microsoft and McKinsey predict that roughly 50 percent of paid work could be automated by 2030. A non-trivial disruption of this scale holds massive opportunity, but it also presents new challenges. On the positive side, workers will spend more time on creative work as repetitive or mundane tasks are automated. On the negative, AI adoption could deepen inequality since automation disproportionally affects the manufacturing and service sectors.
AI creates value, increases productivity, and enhances consumer experience. New machine learning now enables AI to intermediate functions with broad application from customer service to sales and marketing, data analysis, managing schedules and creating reports to people management. The benefits are clear. However, as businesses continue to integrate AI technologies into production processes, they will have new potential risks to consider.
The Macro View
The impact of AI on work and the global workforce are hard to quantify. Some with bullish positions on AI see the disruption as economically comparable to a new industrial revolution. Others predict a slower, more organic process of integration where AI gradually contributes to work processes, augmenting human capabilities without replacing human workers. Beyond speculation, available estimates suggest that AI will constitute a significant share of global economic activity.
91 percent of Fortune-1000 businesses are increasing investment in data and AI. 26 percent are already using AI in widespread production. In the public sphere, 75 percent of governments report that they have, or are in the process of implementing at least one enterprise-wide automated process. Given this widespread adoption and considering planned AI investment it’s fair to say most workers will encounter some form of AI in the workplace, and that most will be required to interact with it.
The foremost fear surrounding AI is the mass loss of jobs and work for global communities. According to most studies of AI currently in action, this fear is overstated. AI is generally implemented as an augmentation device for existing human workers. It rarely completely replaces the role of an existing employee, rather it enhances that employee’s capabilities, through enhanced data analysis.
A second and more pressing issue is AI bias. Bias in AI comes from data analysis and pattern recognition. AI can perpetuate and even augment existing data biases. Here, human elements are essential. Employees working alongside AI should have the ethical training to recognize and correct biases; acting as “emotional circuit breakers.”
Bias manifests itself in several ways, through decision learning that relies on biased and historical training data that informs future decision making and also, through sampling which may over or underrepresent groups in collection and analysis.
State of Play
We are years out from anything like AI ubiquity in the workplace. Expert recommendations on AI implementation focus on a longer-term integration of AI applications in the workplace, backstopped by consistent alignment with organizational culture. Organizations should assess AI’s potential for practical application within specific workstreams to start and build from those proofs of concept as they scale up.
For those on the cusp, the data suggests that larger AI projects with overly broad goals tend to fail fast. Targeted, more bespoke projects and initiatives with specific goals return significant benefits. Business and government agencies are following the augmentation principle. They focus on AI projects that improve products, efficiency, and consumer experience. Public sector organizations for example, are deploying AI to aide in climate and economic analysis, trade surveillance, research, and fraud detection.
The Look Ahead
Market-driven skills development is critical.
Organizations are struggling to secure the talent needed to deploy and operate AI solutions, placing new pressures and presenting opportunities to develop internal capability across operations in both technical and non-technical roles, such as department managers and directors, corporate services, procurement, and contracting.
Education, from K-12 to post-secondary will have to adapt curriculum, modes of learning and engagement to accommodate the future growth of AI in the workplace. As both public and private sector institutions integrate AI into workplace processes, individuals entering the workforce will need to understand AI, both its issues and its potentials.