Thought of the Week

Will the emerging role of AI take your job?

June 6, 2024 4 Minute Read

Artificial Intelligence in business is not new, with many forms of AI and Machine Learning already being widely used. Still the launch of ChatGPT, which reached 100 million active users in two months, has made AI tangible for the consumer. An increase in the personal use and familiarity of AI applications will help accelerate demand for AI in a business context.

As with past technological innovation, AI can profoundly impact where and how people work. This could be through the direct automation of roles, using technology to improve efficiency, or indeed through more fundamental changes either to the type of jobs or the places of employment. This will feed through implications for the wider economy, with estimates suggesting it will generate a global GDP growth of 7% (c.$7 trillion) and an uplift in productivity growth due to generative AI.

Still there are fears that AI will lead to massive job losses as repetitive low skilled or administration roles are lost to automation. And, in contrast to earlier automation models, generative AI has the ability to perform more advanced cognitive work. According to Oxford Economics, this could result in higher skilled jobs being more exposed than in the past. One study suggested around 30% of UK jobs could be automated over a 20 year period. However, historically, many technological advancements have created far more jobs than have been lost. Between 2001 to 2015, automation and technological developments contributed to the loss of 800,000 jobs, but created nearly 3.5 million – a net increase of +c.2.7 million jobs and +£140 billion for the UK economy.

However, using AI appears just as likely to broaden capabilities rather than displacing jobs. It could save time for employees, allowing them to focus on scaling up and strengthening their offering. A key challenge will be competition as rival companies decide the pace at which they invest in and capitalise on the opportunity. Industries that could substantially benefit from AI (e.g., the healthcare sector) might be motivated to take early opportunities. Similarly, companies paying higher wages could be more incentivised to automate work or have increased capacity to finance automation or other AI applications.

But for now, businesses applications and usage remain embryonic. While many organisations plan to use AI, it has yet to materialise. According to McKinsey & Company, while 90% of businesses plan to use both Machine Learning and generative AI regularly, only 35% were using Machine Learning and just 20% generative AI.

It is possible the risks around AI are currently too great. Adopters face risks around inaccuracy , cyber security, risk to user’s intellectual property , and ethical concerns. AI misinformation has been noted as a severe threat in the latest World Economic Forum Global Risks Perception Survey. Even with the best intention, AI can be poorly used. To prevent this, AI models should have a clear purpose and predefined output, should be trained using high-quality, diverse and balanced data, and should be rigorously tested to best prevent hallucinations and biases.

Legislation is necessary to reduce the macro risk around AI, but it still needs to be established globally. The EU Artificial Intelligence Act is likely to be the first. And the UK’s Government-backed AI Safety Institute is also at the forefront of these conversations. Risks aside, growth may be curtailed by lack of resources, especially microchips and electricity.

But we are positive that the applications and opportunities of AI will prevail. Once all aspects of generative AI are understood and trusted, we could see it truly embedded in everyday practice. We have already seen automation through technological development, but the extent to which generative AI will disrupt the jobs market is still unclear, and the expected impacts are nuanced and dependent on the type of work. Generative AI is likely to extend our capacity to automate more cognitive tasks, but ultimately, automation is not new, and it intends to achieve easier, streamlined processes. Instead of replacing labour, it could free it for higher-value work. Humans still have a role to play, not least in approving and justifying AI decision-making with regards to its accuracy and ethics.

Echoing the thoughts of Economist and Professor, Richard Baldwin: