News & blog

Read the latest news and blogs from The Data Lab and Scotland’s data science community.

Human + Machine: combatting myths

Guest blog from Megan Hughes, Artificial Intelligence & Social Media Analytics Lead at Accenture

Artificial Intelligence will be inarguably the most transformative technology of the information age and perhaps for this very reason, there are lots of questions being raised. On the one hand AI stands to fundamentally transform the economy and change the way we work. On the other hand, there’s a plethora of assumptions, fears, hopes and predictions about AI that can be challenging to understand and to navigate.

Much of the dialogue around AI has become specifically focused on individual technologies and their application. Because of this we risk missing the bigger picture, but more importantly we risk losing focus on the incredible benefits Artificial Intelligence technologies have to offer! We need to think beyond the office walls, cast aside the negativity and ask what AI can do holistically, when implemented correctly, for both current & future generations.


Whilst contemplating the art of the possible, it is difficult to ignore the various statistics we are faced with.

Gartner, a leading technology research organisation, predicts that AI augmentation will generate a fantastic $2.9 trillion in business value and recover 6.2 billion hours of worker productivity in 2021 and that there will be an important net gain of half a million more jobs globally!

However, according to the recent OECD Study, Automation, skills use and training, focusing on the risk of automation and its interaction with skills development and deployment in the workplace: close to one in every two jobs could be significantly affected by automation based on the tasks they involve; but the degree of risk varies.

The result of such diverse statistics and the conflicting views they generate is that there are both many myths and a lot of confusion around the topic of Artificial Intelligence technologies which I aim to disperse!

Combatting the myths

The machines are not coming for us!

Firstly, the fact is that the point of singularity, where robots are as intelligent as humans, is far removed from current capabilities. Secondly, total automation of current processes (and roles) using AI, in any case, holds some value but is far from exploiting the full potential of AI. Lastly, going forward into the future, the value of AI is in its power to augment human capabilities – a symbiosis of man and machine, rather than a case of ‘us and them’.

In November 2018 in the UK, there were 753k jobs available and 1.38M individuals actively looking for jobs.

Over the next decade, an estimated 3.4 million workers will be needed to fill manufacturing jobs alone because of Baby Boomers’ retirement and economic growth. 20% of these are likely to be unfilled due to a shortage of workers with the required skills

Over the next ten years, the IT sector will experience the greatest job surge of all. According to, there will be an estimated 1 million more computing jobs than applicants who can fill them by 2020!

Although AI can be deployed to automate certain functions, the technology’s greater potential is in collaborative human-machine teams and its ability to create new jobs.

The Skills Gap

In essence, when envisaging the implementation of Artificial Intelligence solutions humans should do what they do best (creativity, improvisation, dexterity, judging, social and leadership) and machines should do what they do best (speed, accuracy, repetition, predictive capabilities and scalability). The key is unlocking the potential across human-machine teams.

There will be new skills developed through the human-machine symbiosis in what we call the missing middle (pictured below – please read Human + Machine by James Wilson & Paul Daugherty for further information)

On the left side, humans train machines to perform tasks, they explain the machine outcomes, and they sustain the machines in a responsible manner.

On the right side, machines embody physical attributes that essentially extend a person’s capabilities, they interact with humans at scale using novel interfaces and they amplify human insight and intuition by leveraging data and analytics.

It is key to note that none of the skill types, or roles that embody these skills are highlighted in current occupation studies as they simply do not exist (yet) in today’s workplaces.

There is also a difference between individuals who do AI (which does not necessarily involve knowing how to code a Machine Learning application e.g. Knowing how best to ask questions of AI, across varying levels of abstraction to get the insights required) and those who use AI (which really is a broad spectrum e.g. AI in production, customer services, human resources, sales, marketing – the list really is endless).

Thus, the real issue is not actually artificial intelligence (or robots!) stealing jobs or replacing individuals: it’s the potential that AI generates and the resulting skills gap that needs to be addressed.

Time for something new

Finally, the idea that automating and using the same approaches which we have been using for years will continue to work is fundamentally incorrect: mechanistic automation will not continue to yield the kind of efficiency and productivity that businesses require – and certainly not the kind of performance gains those same companies desire.

As businesses deploy AI systems—spanning from machine learning to computer vision to deep learning—some firms will continue to see modest productivity gains over the short run, but those results will eventually stall unless business processes are fundamentally changed (reimagined) alongside.

Ethics, for instance, must be integrated across all AI solutions from the very beginning and not tagged on as an afterthought. Responsible or ethical AI is becoming increasingly discussed, but it is incredibly important to ensure that this topic becomes more than just a conversation. Steps must be taken to build ethics into all implemented solutions and avoid for example any unconscious system and data bias (such as by using Fairness tools).

In conclusion

We must work collaboratively to ensure a sound global public environment that aims to enable and encourage investment in the development and deployment of new AI technologies globally.

Education about the benefits of AI coupled with collaboration across industry, government, and the public on the social, legal, and ethical considerations of AI will converge to ensure a proactive policy framework that really unleashes the full power of AI.

The fact of the matter is, in terms of AI, there really is no defined finish line. It is for us to shape this together, let’s make headway!


Please reach out to me (on Twitter: @megan_j_hughes or via email: for any questions or comments on the above article, the solutioning/implementation of Artificial Intelligence solutions or how to address & implement Responsible / Ethical AI (e.g. fairness tools or ethical canvasses).

Share this story:

Share on facebook
Share on twitter
Share on linkedin
Share on email