On an overcast weekend in March 2017, Bondi beachgoers had the opportunity to get their skin checked for melanomas. But they weren’t attended to by a dermatologist – IBM Watson was put to the task, taking photos, zooming in on anything it thought worrying, and performing a diagnosis.
Watson was taught to recognise three types of skin cancer and 12 benign disease groups. Like a human specialist, its accuracy in detecting melanoma improves as it performs more scans. Research indicates it’s already achieving close to 95% accuracy.
The average dermatologist has accuracy levels of 80%.
If you think artificial intelligence and machine learning won't affect your employment, think again. AGSM Fellow Patrick Sharry says while most people believe automation means robots in manufacturing, they don't realise the impact automation is already having in professional services.
‘It’s already happening in law, for example. The big firms have eDiscovery for litigation and predictive coding for contract review and they’re automating the laborious process of finding precedents,’ he explains. ‘The same thing is happening across all professions. Accounting as a human function will almost disappear.’
Without repetitive tasks, how will we learn?
The rise of AI has been embraced as an opportunity to relinquish the mundane, boring tasks junior staff used to perform, allowing them to step up to more interesting work sooner.
However, without going through those repetitive and fundamental steps, will professionals still be able to gain the knowledge and context they need to do strategic advisory work?
‘If you think about the nature of an expert, which Niels Bohr defined as “someone who’s made all the mistakes that can be made in a very narrow field,” it’s clear we need to do some of the boring work to learn from our mistakes,’ emphasises Patrick.
We’ve been here before
Technology has been disrupting jobs for decades now. The movie Hidden Figures reminded us we once depended on human computers for complex number crunching. Vast typing pools were replaced by desktop software; bank teller numbers were decimated by ATMs.
But the pace of change is accelerating and it has implications for every market sector.
The Institute for the Future’s Future Work Skills 2020 report predicts that within 10 years, ‘new smart machines will enter offices, factories and homes in numbers we have never seen before. They will become integral to production, teaching, combat, medicine, security, and virtually every domain of our lives.’
Even some of the things we believe to be uniquely human, such as creative arts, can be replicated by machines. A deep-learning machine has used algorithmic processes to create an entirely new Bach concerto, and write pop songs in the style of The Beatles.
From knowledge to co-creation
In their book, Humility Is the New Smart: Rethinking Human Excellence in the Smart Machine Age, Edward Hess and Katherine Ludwig write about the ‘NewSmart attitudes and behaviours’ we need to adapt to the Smart Machine Age.
‘In many professions, the traditional model is “I deserve to be at the top because I know things”,’ explains Patrick. ‘But now machines can know more and analyse faster than any human.’
Hess and Ludwig say that means focusing on the things machines don’t do well – critical, creative and innovative thinking, and genuine engagement with others. They describe four key behaviours, which include quieting your ego and reflective listening.
Some human functions may be replaced by automated systems but others will be augmented – and this is the opportunity. We can extend our own capabilities by working in ‘human-machine’ collaboration.
KPMG has dubbed this ‘the hybrid workforce’, where people manage machines, and machines manage people. We’ll learn to co-create with digital agents (virtual assistants, like Amazon’s Alexa) and cognitive computers. We’ll make fewer decisions based on gut instinct, and depend more on the ever-increasing amount of data available. Our workplaces may have fewer people, but they could be incredibly productive.
Just imagine the impact if IBM Watson could find more melanomas sooner, reach more people in remote areas and sift out the false alarms. Dermatologists could then focus on diagnosing and treating the dangerous lesions. That’s human-machine co-working with real impact.
‘Ultimately, it’s all about what we can do with the machine,’ concludes Patrick.
 Training Watson to Help Detect Melanomas Earlier and Faster, Adrian Bowling, March 2017, IBM.
 Future Work Skills 2020, Institute for the Future, for the University of Phoenix Research Institute.
 Deep-Learning Machine Listens to Bach, Then Writes Its Own Music in the Same Style, MIT Technology Review, December 2016.
4 People and robotics – the hybrid workforce, KPMG, 2017.