ASKING today which industries are going to be most impacted by artificial intelligence (AI) in the coming years and decades is much like someone in 1980s asking which industries are going to be most impacted by computer chips and computer software.
In 1980s, it was possible to see that computers were going to revolutionise retail and medicine, but not to forecast the advent of Amazon or Taobao. In fact, the prediction of AI’s future influence is probably even more difficult than was the case for earlier technologies, due to the exponential nature of change in an era of unprecedented connectivity.
What is clear is that AI and its associated technologies will continue to advance. In Southeast Asia alone, the adoption of AI technologies stood at 14 per cent in 2018, a marked increase from 8 per cent in the previous year.
The use of AI will only trend upwards as organisations turn to this technology to plug the skills gap created by the digital transformation of traditional industries. This is set against the backdrop of a potential talent deficit of 47 million workers across Asia Pacific by 2030.
As AI becomes mainstream, today’s AI capabilities will expand to address more general uses, otherwise known as Artificial General Intelligence (AGI). Once this happens, there will be fewer tasks that humans are able to perform more efficiently or economically than can machines. The unique economic value of humans will then be restricted to special cases such as psychological therapy, entertainment and arts, or roles that require human engagement or interaction to deliver value.
Before we get there, we must acknowledge that AI will not generally take over the human workforce all at once. Instead, it will start by prompting businesses to reorganise their structures to best leverage the strengths AI has to offer. In doing so, portions of people’s jobs will be reallocated to AI, giving way to the creation of new roles and requiring humans to upskill themselves accordingly.
For example, AI will first take on more routine accounting work such as calculations and backend processes, similarly to intelligent adaptive Excel macros, while allowing accountants to handle strategic decision-making and customer interactions. Only with the emergence of AGI over time will we see AI taking over these roles and by then, only a handful of humans will be needed to oversee the running of an organisation’s entire finance function.
When this happens, we would be living in the reality of the hyper-intelligent workplace, where humans will simply provide mission-critical thinking, while incorporating the intuitive skill sets of AI and robots for increased productivity.
A key thing to note is that AI will not instantaneously eliminate jobs or job functions. Using the case of biological lab testing as an example, the equipment which is already computerised in the main will become more automated and intuitive. Lab robots will then start bridging the gaps between the lab equipment – be it carrying materials from one machine to the other or keeping each area of equipment informed about their respective test subjects, parameters and results.
Thus the role of a lab technician would be gradually replaced by an integrated lab robot system that takes verbal instructions directly from the scientist in charge of the lab. Eventually, the scientist’s role will mainly be to rubber-stamp the AI’s suggestions, knowing that what the scientist needs from the AI has been effectively executed.
Of course, this is not to say that there will be a widespread of automation within all industries. While generic graphic designs can now be done with AI design programs, the conceptualisation of deeply creative new images and visual themes will remain largely in the domain of humans, especially on highly subjective matters that require a highly nuanced understanding of consumer taste and preference.
In general, one may say that among the last job roles to be eliminated will be those relying heavily on novel creativity, strategic decision-making, critical thinking and physical manipulations in life-or-death situations. Yet, among these roles, the routine and labour-intensive tasks will most likely be automated first as that is where most cost savings will be obtained.
To put it into context – there is more money to be saved by automating car driving than helicopter piloting, and more to be saved by automating medical research than particle physics research.
In my view, the reduction of the need for humans to work for a living should be massively positive. In an ideal world, this would free up people’s time to pursue more social, artistic, intellectual and spiritual endeavours. Imagine living in a society where, through the work achieved by technology, the psychology of defining one’s importance, status and identity no longer relies on one’s career or income.
While that is almost sure to be a rocky road given the realities of geopolitical issues and income inequality, projects such as the decentralised AI movement led by SingularityNET aims to smoothe the path towards it in a democratic and participatory way.
Contributed by BEN GOERTZEL, CEO of SingularityNET