I recently spent some time tinkering with deep learning algorithms that combine paintings from famous artists with my own photographs to create new images in the style of the artist. I’ve been quite taken with some of the results. I like what a modern-day Van Gogh has done with Brighton pier, his distinctive brush strokes proving relatively easy for the algorithm to convincingly mimic. Other experiments show just how far AI has to go before it can replace human creative ability; Lichtenstein’s Ben-Day dots tended to give portrait subjects a dose of something alarmingly contagious.
All in all, I’ve concluded that the job of famous artist isn’t at risk just yet. But of course, many other roles are at greater risk of AI-induced displacement, that is, if such displacement hasn’t already occurred. And it got me thinking about how we’re responding to the risks and opportunities presented by the technology.
In the cracking “AI Super-Powers, China, Silicon Valley and the New World Order” (2018) Kai-Fu Lee illustrates how displacement risk varies across occupations and is determined by the extent to which the number of tasks within those roles can be performed or bettered by technology (Figure 1a). Figure 1b shows how the displacement might evolve, and meatime gets us thinking about how roles in the Augmentation Zone might leverage technology to help complete the tasks involved. Roles that require high levels of social interaction and creativity or strategic planning will remain further from the reaches of AI for longest, whilst those involving limited human interaction, or in highly structured physical and non-physical environments are most at risk.
It’s important to note that AI doesn’t differentiate based on traditional measures of skill level for a given task. For example, a recent report in the journal Nature showed how Google has developed systems that can now detect breast cancer in mammograms more accurately than skilled human radiologists. This is in contrast with very limited progress that technology has made in automating many service industry tasks such as cleaning hotel rooms or folding laundry*.
While many individuals may not be concerned with an imminent risk, those in the Augmentation Zone and beyond must be willing to embrace AI technology in the near term for improving task efficiency if they are to remain competitive and relevant in their chosen industries.
Lee goes on to talk about how, in 2019 via one of his famous letters to CEOs, BlackRock’s Larry Fink issued a call for businesses to consider what impact certain factors will have on jobs and society in future, technology being a key one. Fink views the effect of technology on jobs as a driver of public anxiety, frustration and deteriorating confidence. In the letter, he asks public and private companies to take a more proactive role in “providing lasting solutions…to address pressing social and economic issues.”, particularly in view of government’s failure to do so.
Governments have been slow. Like the rest of us, they are battling to keep up with the dizzying pace of innovation. It is difficult to accurately predict the direction that technology – particularly the AI-powered variety – will take, making policy setting increasingly difficult and risky. Faced with this uncertainty, governments have become too passive. But as AI continues to transform the landscape for employers, employees and broader society, government intervention will undoubtedly be needed. They must urgently work to ensure they get closer to the leading edge so they’re better able to deal with the implications of the technology. The potential scale of the intervention needed should not be underestimated and some have at least started to consider the possibilities. Measures like Universal Basic and General Minimum Incomes, or stipends that reward individuals for taking on roles in care, community service and education, will likely become necessary to guarantee individual financial security in the face of large-scale job displacement.
But what should businesses be doing? Well, as the Displacement Zone reaches further across the spectrum of roles, so it becomes increasingly important for organisations to take on responsibilities beyond current fiduciary boundaries. Fink talks about profitable companies “effectively serving all of [their] stakeholders over time – not only shareholders, but also employees, customers, and communities”. Lee suggests this may take the form of “reimagin[ing] and reinvigorating corporate social responsibility, impact investing and social entrpreneurship”.
AI will offer unprecedented opportunities to increase efficiency within these companies, thereby reducing cost. It will be extremely tempting, no doubt irresistibly so in some cases, to take these efficiencies straight to the bottom line. But it’s precisely here where smart leadership can respond to Fink’s call, enabling transformation of customer and employee experience in very practical ways.
At organisational front-lines for example, this could be about investing in better resolution of customer issues by allowing employees more time to provide remedies, continuing the trend of moving away from pure-productivity driven operations to deliver more “Wow” moments for customers. Similarly, it could be about giving those same employees more latitude to investigate and report on issue root causes, increasing their responsibilities, giving them greater ownership and involvement in solving operational problems so they don’t recur.
In sales environments, it might take the form of investment in more powerful tools enabling employees to deliver knock-out sales experiences. Many businesses remain reliant on human interactions at the sales front-line, where conversion rates can be much higher than the digital equivalent, and my Lichenstein portraits serve as a reminder that it will be some time before AI is able to deliver the necessary human-touch here.
Elsewhere, as the demand for human resource to carry out traditional tasks reduces, businesses might use the opportunity to further increase the flexibility of working patterns for employees, allowing them to choose shifts and hours that better match other personal or family commitments, in line with seasonal demand. In my experience, such flexibility is a very effective driver of employee engagement.
Such examples are relatively low-level, and don’t begin to consider some of the wider societal- and community-level contributions that will be needed, from large employers in particular. However, they demonstrate that by leveraging at least some of the AI bounty to provide better services and customer outcomes, it’s possible to imagine a future in which these improvements positively impact an employee’s sense of purpose and contribution to organisational goals, leading to increased loyalty and engagement with the organisation.
So, while my painting experiments show that AI still has much ground to cover before Banksy or Damien Hirst need worry, the rest of us must face in to the reality that AI-induced change in the workplace will become more and more routine. We must be ready to take the opportunities the technology creates, and be prepared to take ourselves and our businesses in some fresh, and potentially very rewarding directions.
* – That said, progress is being made: https://youtu.be/-V2nHp7n7k0. The phenomenon that seemingly simple tasks can often be harder to replicate through technology than cerebrally difficult tasks was observed back in the 80’s when AI pioneer Hans Moravec first stated that “it is comparatively easy to make computers exhibit adult level performance on intelligence tests or playing checkers, and difficult or impossible to give them the skills of a one-year-old when it comes to perception and mobility”. Moravec’s Paradox remains as true today as it did then, despite significant progress on both types of task.