Kathryn Cave on April 17 2017
Middle management gets a very bad reputation. Italian designer Massimo Vignelli succinctly described it as a “disease” while Steve Jobs took this whole tier of the workplace apart in a bit more detail:
“Companies, as they grow to become multi-billion-dollar entities, somehow lose their vision. They insert lots of layers of middle management between the people running the company and the people doing the work. They no longer have an inherent feel or a passion about the products. The creative people, who are the ones who care passionately, have to persuade five layers of management to do what they know is the right thing to do.”
This whole strata of the workplace may have been a political minefield for decades but it could still be ripe for automation. At the end of last year the Guardian reported that the world’s largest hedge fund, Bridgewater Associates, was “building a piece of software to automate the day-to-day management of the firm, including hiring, firing and other strategic decision-making”. While software testing consultancy, Piccadilly Group has similar ambitions but for enterprise transformation projects.
“Major banks have generally spun up teams of up to a thousand to achieve [their transformation] goals,” explains CTO, Adam Smith when I meet him and his CEO Dan Hooper in London. “This can take 30 people to manage. And it is not unusual for 50 or 60 PowerPoint slides to be distributed on a daily basis [to explain how the project is progressing].”
Piccadilly Group has interesting insight into this field because it is mostly drafted in by large banks to get drifting big budget transformation projects back on track. Now, in order to better this process, it is building an AI platform which it hopes will begin to strip out the whole unhelpful middle level of management in the next two years.
“The end goal is to replace a lot of mid-tier roles,” explains Hooper. “We’re getting rid of the politicians.” There is a whole industry chasing people down, he adds. Instead this aims to “directly connect someone who does code with the CEO. It is consultant killer.”
Smith is keen to stress that all this is possible in this particular area because “software testing is super well defined”. Few of the AI startups are looking into the real world, he says. They tend to be very broad and are trying to tackle expansive problems. “Narrow AI” – or clearly defined AI – on the other hand is perfect for industry because its aims are very clear. Technologically speaking “we’re using Natural Language Processing [and other AI techniques] in a fairly standard way,” he adds “what is interesting is we’re applying it to the enterprise”
The demo I’m shown is impressive but more for its potential than what it delivers right now. The interface is a fairly typical chatbot that assimilates in data from a variety of disparate systems to show where the project is at in terms of predefined goals. This allows it to provide simple steps on what needs to be done, by when, to keep the project on track and to deliver visibility on where any blockages lie. So, instead of a fraught ring round it might tell you straight out that six developers are off sick and three haven’t logged into the project this week.
In the longer term this should enable a CIO to pull instant graphics, graphics and timelines across the entire project but at present Piccadilly Group mostly uses the data visualisation element of the platform internally. Smith estimates it will take two years to train the AI and prediction engine to the point where it will be able to replace people.
Exactly how many jobs will be automated out over the next few years is difficult to predict and there are several different estimates doing the rounds. Yet what is striking is that it is not just manual roles that are threatened but also a raft of highly paid professional positions. This means whether this particular platform proves effective or not, what makes it interesting, is it is a deliberate bid to replace relatively senior people in an area where tasks can be defined well enough to make it possible.