Amidst all of the predictions of AI technologies taking our jobs en masse, I’ve always tried to project a slightly more nuanced, and hopefully realistic picture both of the state of AI today and its potential impact upon the workplace.

Whilst I do very much believe that AI and other automated technologies will be very powerful, I think that they will yield the best results when working alongside humans, with each performing tasks that are best suited to them.

It’s a perspective that’s shared by Accenture’s Paul Daugherty and James Wilson in their latest book Human + Machine.  In it they argue that the best results come when companies recraft processes specifically for AI rather than trying to bolt them onto legacy systems.  It’s a process they refer to as the ‘third wave’ that sees AI complementing rather than replacing humans.

Third wave organizations

They outline five key principles that organizations already behaving in this way exhibit and live by:

1. Mindset – the best organizations have a radically different approach towards business that sees human employees improving AI, and AI in turn improving the performance of human employees.  A crucial first step is to develop the potential of employees to apply automation to their various routine tasks.
2. Experimentation – AI remains largely untested in many organizations, so successful implementation will require a constant search for ways in which processes can be improved by the technology, and an experimental mindset to try things out.  This is charting new territory so organizations will have to learn as they go, so the ability to create tests to derive business processes that work will be vital.
3. Leadership – the best organizations make a specific commitment towards the responsible use of AI from the start.  The ethical, social, moral and legal implications of AI have been well investigated.  Therefore, executives must promote AI that is explainable, transparent accountable and free from biases.
4. Data – most AI technologies today run on data, so it’s vital that organizations get their data house in order.  The accumulation and preparation of data is one of the biggest challenges for the successful deployment of AI systems today.  The best organizations have data that flows freely across departmental silos.
5. Skills – to work effectively with AI will require what the authors refer to as ‘fusion skills’.  The next wave of AI technologies in the workplace will require humans to design, develop and train AI systems, as well as collaborating effectively with them.

The book provides a practical roadmap for organizations that is based around these five principles.  I’ve written before about research from Accenture on how to blend man and machine to get the best out of AI technology, and the book is very much an extension of this line of thought.  For those looking to experiment with AI in their workplace, it’s well worth a read.