The Impact of AI on Organisation Design

The Luddites world tour 2023

Arthur C Clarke once observed,

‘Any sufficiently advanced technology is indistinguishable from magic.’

The latest form of witchcraft is artificial intelligence, and the debate is rather predictably focused on the impact on jobs.

‘Predictably’ because we have been here before – many times.  For centuries disruptive technologies have been greeted with warnings of mass unemployment.  And each time they have been proved wrong.  It’s called the Luddite fallacy, after the 19th century textile workers who destroyed the automated looms that had robbed them of their jobs.  Although undoubtedly painful for the displaced workers, the wider population benefited greatly – cheaper clothes, the creation of new jobs in other parts of the economy and a consequential general increase in living standards.

This pattern has repeated time and again since the beginning of the Industrial Revolution.  It seems that the story of human progress is one in which we seek ever more inventive ways of using fewer people to acquire the things we desire.  And, in so doing, we have released people to pursue activities of even greater economic, societal and personal value (Maslow’s hierarchy of needs anyone?).  Consider agriculture.  Today it accounts for 1-4% of the labour force in developed countries.  Before 1800 it was over 60%.  Hands up who wants to give up their smartphone, flat screen TV and electric car so that we can return 56% of workers to tilling the earth? 

Reasons to be cheerful

So, is history about to repeat itself with AI?  The optimists argue that AI will:

• create more jobs than it eliminates, as has always happened and as is predicted by current research

• make jobs more skilled and interesting, and provide employees more autonomy and control, not less

• enhance job security, not weaken it

• have a positive impact on physical and mental well-being in the workplace, automating the dull parts, leaving more time for the interesting parts

• lead to an estimated 26% increase in global GDP by 2030 (World Economic Forum).

It was technological advance that overcame the Malthusian Trap, lifted us out of poverty and raised living standards.  And in the AI age, where would we now be without search engines, spell-checkers, digital assistants, or Netflix recommendations?

Reasons to be cautious

Economist William Nordhaus believes that we are some way from an economic ‘singularity’, the point at which superintelligent machines render humans economically superfluous, enabled by artificial general intelligence (AGI) which will outperform humans in physical and cognitive tasks.

On the plus side, if humans no longer need to be economically active (Maslow’s hierarchy of needs again anyone?), we have a real chance of ending inequality.  On the downside, we might simply be sleepwalking towards Aldous Huxley’s dystopian Brave New World.

But for now, perhaps consider that:

• AI will increasingly displace low-skilled, repetitive work, but will expand high-skilled work (for example, data scientists working in bioinformatics)

• AI may create a false sense of security – that it always produces correct results.  Amazon had to abandon an AI recruiting tool when it was found to be biased against women.  The system had been ‘trained’ to look for patterns in CVs, most of which came from men.  A bias towards male traits was built in from the outset – the system taught itself that male characteristics were preferable

• the use of AI surveillance tools could have a huge impact on trust, especially regarding privacy at work

• tools designed to increase efficiency, improve consistency and reduce waste, might do so at the expense of personal autonomy (telesales workers disciplined by their supervisors because the AI detected them logging on 14 seconds late – real world example!)

• an AI never sleeps, so people might feel the pressure to be ‘always on’, which has a negative impact on job performance

• if we become dependent on a few tech behemoths (who enjoy both monopoly and monopsony power), the interests of humanity are not well served – for once the Luddite fallacy might turn out to be true.

Now is the time to adjust

AI is already affecting your business, for good or ill.  And if you don’t believe that, we have a bridge to sell you.  Whereas much is being said and written about the need for regulation, the impact on jobs, the promise of a better tomorrow, little is being said about the impact on organisation design.  So here goes.  Some significant adjustments to ponder:

• As ‘traditional’ roles are eliminated and new ones created, organisations will need to adapt more quickly – make strategies more fluid and emergent, moulding rigid organisational structures into something more agile, and investing significantly in upskilling and reskilling to give people new skills.

• AI cannot (yet) replicate the soft skills needed to manage employee relations, to inspire achievement, or to reach consensus.  So the value of creativity, leadership, coaching and collaboration will likely be enhanced.  This, in turn, implies vesting more autonomy in one’s staff and decentralising authority and decision-making power.

• The very nature of management will change because AI will disintermediate traditional forms of coordination and control.  Managers should expect to demonstrate more judgement-related skills, such as creative thinking, experimentation and data interpretation.

• Managerial spans should widen substantially as AI increasingly reduces the requirement for direct supervision. 

• Organisations are likely to become flatter as employees will more readily gain direct access to critical information and decision-making tools, leading to quicker decisions and greater innovation – your organisation is going to move faster.

• Middle managers and administrative support functions will be most affected, shrinking the distance between producer and executive and increasing the ratio of direct to indirect staff.

The most obvious use for AI is to automate work and to simplify and improve decision-making.  Processes, responsibilities, organisational structures and culture will need to change.  Organisations where people and AI work together will garner a great advantage.  The concern is less that AI is coming for your job, but that someone wielding AI might be!