How can we prepare for the AI revolution?

Artificial intelligence (AI) promises social and economic change on a par with the industrial revolution.  However, this profound change will be squeezed into a single generation, rather than the several generations over which society was able to adapt to the impacts of the first industrial revolution.

We have barely begun to see the effects of AI, and knowing the full extent of its future impacts is impossible. However, we highlight four broad areas of impact that we believe will prove particularly important:

  • The potential for huge economic benefits
  • Labour force disruption and social adjustment
  • Concentration of power and regulatory challenges
  • Public institutions and fiscal health

Over time, we expect governments to respond to the challenges of AI. The adjustment process is likely to be bumpy, as the opportunities brought about by AI might ripen faster than governments’ ability to address their downsides.

The potential for huge economic benefits

The range of consultants and industry experts who have explored the topic thus far are almost unanimous in the view that AI will deliver significant economic benefits. According to some estimates, we could see an additional $16 trillion worth of output from AI alone by 2030, helped by significant rises (between 11% and 40%) in labour productivity.

If these numbers play out, the implications will reach far beyond niche technology sectors. The McKinsey Global Institute predicts 70% of companies will adopt at least one form of AI by 2030[1] and our anecdotal discussions suggest the topic has been discussed in most boardrooms already.

Labour force disruption and social adjustment

Technology that is able to form decisions autonomously could revolutionise jobs that have relied on judgement, knowledge and insight. The result could be significant upheaval in the workforce.

However, the issue is less one of permanent job losses and more of the need for the labour force to retool with technology skills and to adapt to the different roles people will play in a newly-networked economy. This will be vital to the ability of companies to remain competitive and enjoy future success.  

The adoption of AI into the labour force is likely to start with existing leaders augmenting human experts with AI techniques. The balance of decision-making should move to the latter over time as their effectiveness is proven.

Providing the training and support to societies to allow them to make that transition smoothly, and the safety net to those who cannot do so, will be vital to mitigating social unrest. With societies becoming increasingly disenfranchised, providing that safety net could be the key to protecting the role of public institutions.

Concentration of power: how companies should respond

Artificial intelligence is likely to create self-reinforcing positions of dominance for incumbent leaders. The largest companies accumulate more data, providing a competitive advantage. If exploited wisely, it will allow them to develop more powerful insights and actions, bolstering their scale and strengthening their competitive positions.

For the foreseeable future, technology leaders look likely to maintain their dominance, with corporate power concentrated in the hands of a few. This risks leading to unhealthy monopolistic behaviour in which social welfare and consumer or worker interests are subordinated to profitability with too few checks or balances on their actions.

This concentration creates headaches for regulators, particularly when charged with overseeing intangible technology sectors where assets are internationally mobile. We expect a growing political focus on limiting the potential power of AI leaders. Over time, a regulatory response is likely, limiting the scope of AI use or requiring information sharing to aid competition. A key focus is likely to be on providing transparency, so that companies are required to explain how algorithmic decisions are reached.

In our view, it will be increasingly important that companies emphasise the benefits their use of data and AI can deliver, rather than focusing on their control of both as a source of competitive advantage.

Protecting public institutions – and tax revenues

Governments are already struggling to grow tax revenues sufficiently to cope with the growing demands of ageing populations. Income is likely to become concentrated in the hands of fewer individuals or companies, creating greater challenges for governments trying to tax them. The shift toward growth in intellectual property-intensive industries has also begun to create fiscal difficulties for cash-strapped governments. Many of these have responded through coordinated efforts (in particular through the G20) to ensure technology companies (who have historically proven adept at minimising tax payments) pay taxes in line with their scale. The EU is considering introducing a 3% tax on revenues in the sector.

Looking forward, more revolutionary tax reform and greater international cooperation is likely to be necessary to address the threat of shortfalls which might otherwise undermine governments’ abilities to provide the services societies expect. It will be more important than ever that companies focus on business models that deliver social benefits and profitability, without relying on aggressive tax strategies to support earnings.

The formula for AI success

Drawing concrete conclusions on the impact of artificial intelligence, which remains in the infancy of its potential impacts, can only be conjecture.  As scrutiny grows, we believe the companies that are most likely to succeed are those that:

  • Plan for structural retooling of their workforces
  • Build competitive advantage through their use of data and AI to deliver social benefits which sustain their license to operate, rather than commercial applications which cannot demonstrate their public value
  • Avoid aggressive and unsustainable tax strategies or regulatory boundary-pushing that could unravel as fiscal intervention toughens

Social challenges are unavoidable and we expect governments and regulators to become increasingly focused on action to address the resulting difficulties. That political response will help sift companies that have prepared for those challenges from those which have not.