In the final year of our research programme, we’ve witnessed what’s widely believed to be a step-change in digital innovation, resulting from advances in generative AI. In some ways it’s a light bulb moment! The first street lighting in Paris for the 1878 Exposition Universelle, or in 1880 the first town to be fully lit by electric lighting in Wabash, United States. But it was a light bulb moment as many entrepreneurs, firms, and public authorities at the time, as they soon realised other innovations could be developed as a result of rolling out electrical power to illuminating our towns and cities.
Recent advances in what is called ‘generative’ AI – given its capacity to create ‘content’ including text, images, and code, based on hundreds of billions of data points – has a long way to go to get anywhere near the significance of innovations in electric lighting, but the launch of ChatGPT by its creators at OpenAI in San Francisco, led to a surge of over one million users within the first week of launch in November 2022, and has been picked up by media outlets globally. Other companies such as Google’s LaMDA conversational dialogue, Midjourney’s text to image, and DeepMind’s Gopher language model, along with others such as Nivida, are all engaged in similar developments, and this is before the widespread application of quantum computing.
Today’s light bulb moment for anyone interested in the future of work, is the fact that unlike robotic process automation (RPA) that claims to free labour to pursue more skilled tasks by ‘taking the robot out of the human’, generative AI challenges the very foundations of the ‘knowledge’ economy.
We’ve framed the world’s education systems around the development and assessment of cognitive abilities. An example being the race to higher education in recent decades, based on an anticipated increase in the demand for knowledge workers in a high tech, digital economy. Using data from the 2018 OECD PISA survey, Anthony Mann, a Senior Policy Analyst at the OECD, found that seven in ten 15-year-olds across the OECD now expect to achieve tertiary qualifications and almost two-thirds (62%) expect to work as a professional or in a managerial role, a proportion that has risen from 53% in 2000.
But what if the value added of a knowledge or digital economy is not in the heads of workers, but captured in algorithms and on ‘smart’ platforms? While there are many kinds of knowledge work that can’t be directly automated, many professional, managerial, education services, can be redesigned to concentrate the productivity of human knowledge in a few hands aided by increasing powerful digital tools. Having said this, technology is not fate.
There are different business models of productivity linked to generative AI. From a cost-reduction and/or war for talent approach, there will be an active commitment to empower those at the top of organisations to extent command and control over all aspects of the business, including greater control over content using smart platforms and applications. However, the further separation between conception and execution this entails, where more of the conception is aided by generative AI runs new dangers for companies given the potential for generative AI to result in error, false data, and flawed decision-making.
If oversight is in the hands of a few with what can be called ‘digital scribes’ being part of a narrow talent pool, companies will be blind to the collective intelligence available to organisations that harness technologies to extend the capabilities of the whole workforce, many of which are more attuned to the daily operations and changing needs of customers. They also serve as an early warning system when there’s a problem, as well as recognising new market opportunities. But this will require a more people-first or human-centred approach, where permission to think is enhanced rather than captured by technology.
It will also need different government approaches as it is no longer viable to work on the premise that new technologies are skills biased, creating more and better jobs that those automated.1
1Phillip Brown, Hugh Lauder and Sin Yi Cheung (2020) The Death of Human Capital? New York: Oxford University Press.