How will automation affect the third sector?

I recently spoke at The Gathering, the annual event for Scotland’s charities and third sector organisations, run by SCVO. Alongside me on the panel were Dr Kendra Briken from the University of Strathclyde and Rhodri Davies from the Charities Aid Foundation. The question we were posed was, ‘How will automation affect the third sector?’ We all came at it from different perspectives and I found myself furiously scribbling notes on my phone as Kendra and Rhodri spoke. Here’s (roughly) what I said. What defines your organisation? Is it the mission? Is it the values? Is it the culture and behaviours? The coming wave of automation is going to make you think very hard about these questions again. Because the choices you make won’t just about technological possibilities, they will be about your strategic priorities. Automation is nothing new. Since the first Australopithecus afarensis sharpened a rock three and a half million years ago, we’ve been applying our understanding of the world to make our lives easier. We are a race of toolmakers and as our tools get more sophisticated, they can take on more of our work for us. In every wave of technological advancement in the relatively recent past, from the agricultural revolution through the industrial, technology has enabled economic growth, which has in itself created more work. But it looks increasingly like this will not be the case this time. General purpose computing can be applied so widely to such a range of tasks that it is hard to see how it doesn’t result in either large scale unemployment or its modern equivalent, the further degradation of high quality jobs into gig economy work. Let me give you some examples: In China, the government’s Made in China initiative promotes the rapid automation of its industries. Some cities, like Dongguan, are even offering grants for companies to invest in machines that displace people. The results have been dramatic. 87,000 workers were replaced in Dongguan between 2014 and 2016. Apple’s largest supplier, Foxconn, replaced 60,000 workers in a single factory with robots. This is physical automation, something we have been familiar with for centuries. But it is newly enabled by the cheap availability of computing power. These robots aren’t just mechanically sophisticated, they are digitally smart. This intelligence translates into automations that will potentially be very relevant to the third sector. The first is what is happening in stores, with systems like Amazon Go. Here, tills are eliminated by a network of cameras placed around the store. People tap their phones on a terminal on the way in, and then whatever they pick up and put in their bag, they are charged for. This may look like it will only be available to high tech stores like amazon or maybe the big supermarkets. But the hardware required is actually very simple and cheap. The likelihood is that Amazon and others will offer such systems as a service to retailers at a relatively low monthly cost. That cost may still be more than a volunteer. But as more and more shops adopt this technology, queuing at a till to make a purchase is going to start to look pretty archaic. Like the 20MB data limit on the wifi, on my train here. Stores using this technology are going to know a huge amount about their customers or supporters. It will also facilitate very rapid pricing and stock taking. The second big physical shift that may be relevant is the advent of autonomous vehicles. This class of automated transport and delivery systems will range from the pavement-roaming Starship drone to the driverless car. And they will likely be here sooner than you think. This drone has already been trialled on the streets of London.

Less anthropomorphic

When we think about automation, it’s natural to think of physical tasks and machines that in some way perhaps resemble us. After all we have been replacing humans with machines in a manufacturing and agricultural context for a few hundred years. But what about automating more cognitive and interactive tasks. Like call centres. Amelia is a virtual member of staff that can interact with people via voice or text. She doesn’t need to be programmed with answers. You feed her information, connecting her to your CRM systems, etc, and she finds answers based on the questions she is asked. She is sensitive to emotion, and tailors her responses accordingly. Amelia can answer every call on the first ring. She can typically answer 80% of queries. And scariest of all, people often prefer dealing with her to dealing with another human — particularly when dealing with potentially embarrassing issues.

In the back office

Amelia is a particularly visible kind of automation, but there is a much less visible kind as well. The automation of document and data processing. An incredible amount of work in most office environments is still devoted to re-keying data, reproducing documents that are very similar to those that have been produced in the past. It has taken much longer than anyone expected, but this friction in day to day operations is slowly being eliminated. The barriers between different systems are being broken down. The production of reports and documents is being automated. This isn’t so much automation as augmentation. Few whole roles will be eliminated by this type of technology. But with it, fewer people can achieve more, meaning you don’t need as many people in total.

Structural change

Automation as a general rule more greatly affects the lower levels of the hierarchy, where work is more often repetitive. As you climb the hierarchy and work becomes more about creativity and adaptation, automation becomes augmentation: a smaller number of people equipped to do more. This has the potential to change the shape of the organisational hierarchy: low down a small number of people augment and maintain a largely automated workforce. Higher up there may actually be more people in those organisations that successfully make the transition to a more automated world. Their creativity and relationships with customers or partners may be the bottleneck that limits growth. Ultimately though, we return to a pyramid at the top.

The response

In the next few years, you can undoubtedly apply capital to reduce your operational costs through automation. Success is a question of good design as much as the right technology, but it’s a realistic prospect and the investment levels required to start to have an impact are actually relatively low. The question comes back to your priorities. Are you here to achieve your mission at all costs? In which case, there’s a strong argument you should be pursuing every solid prospect for automation and augmentation reducing your costs and increasing your efficiency. Or is your charity also a vehicle for wider social good? Is it important that you create employment? Finally, where do volunteers sit in this mix? How much do they actually cost you? And could their work be done, or at least supplemented by machines? This might be the trickiest area to assess.