Last week, I attended a conference hosted in Bengaluru by the firm Automation Anywhere Inc. Automation Anywhere, and its competitors such as UiPath and Blue Prism, are in the business of creating “bots", or discrete pieces of software that can automate a lot of work carried out by office workers today. In a reflection of how hot the area is today, the conference had plenty of marketing pizzazz.
The first targets of these bot makers are jobs that do not have great complexity. Think of an accounts payable clerk in your office for instance. All he or she does today is to complete a two-way or a three-way match of the amount payable against an invoice—with an existing purchase order—or with both the purchase order as well as another document, such as one that records the receipt of goods specified on the invoice, before approving the invoice and initiating payment.
The worker is usually using computer systems such as enterprise resource planning (ERP) systems that automated the paperwork associated with such transactions in a previous wave of automation that took place a couple of decades ago. At that time, it was systems from Germany’s SAP or the US’ Oracle that automated these paper-driven processes to create the sub-class of “Information Technology Enabled Services" or ITeS. Indian outsourcing firms had a fair share both in the implementation and the ITeS administration of such ERP systems.
The first step in setting up a bot is to create a “recorder" that cohabits a worker’s computer. It then learns the basics of the job over a few days, by recording keystrokes as the worker does his or her job, noting which parts of ERP systems are accessed by the worker, and in what sequence, to complete a series of tasks. In a short while, the recorder learns a large proportion of the repetitive tasks a worker performs.
After the recording has been completed, a team of computer programmers converts the recorded actions into an automated script that can run on demand. Interestingly, the amount of programming required to create a bot is quite minimal. Since the bot can now totally replicate what the human has so far been doing, it simply replaces the human.
This is not the job loss we fear from Artificial Intelligence. It is the replacing of manual processes with mechanized ones, much as a tractor replaces the need for multiple farm hands. This last bit of office automation by bots is just the icing on a cake that was baked 20 years ago.
This next step of automation was always inevitable. It would have actually taken place in the early 2000s if the Indian offshoring industry had not risen. All that the rise of Indian IT and ITeS offshoring did was to provide a mezzanine step, since it was cheaper to export the repetitive work to destinations like India or the Philippines than it was to automate them. A mezzanine landing is not meant to be occupied for a long time; traffic must inevitably make its way upward—or downward—on the staircase.
Predictions about the job loss due to new technologies are imprecise, largely because bot-led automation is already here, while Artificial Intelligence growth is on a trajectory that is difficult for most experts to predict. The one prediction that seemed most balanced was from a report called Jobs Lost, Jobs Gained by the McKinsey Global Institute. The report includes many of the “it depends" type of answers that is the wont of consultants, but is nonetheless a fine piece of work. Its prediction? India will lose 56.9 million jobs—or “full-time equivalents"—by 2030.
In what I believe is a misguided attempt to make bots seem more human-like, some of the press coverage on them has taken on the grotesque practice of counting bots alongside human employees, as if these bots were somehow worthy of personhood. For instance, recent press coverage of an IT services provider said its workforce consisted of 17,000 plus employees and 350 bots or “virtual" employees. Another story says that the Indian back office of an Antipodean bank counts 3,500 “virtual" employees among its workforce.
Bots are not “virtual" humans. At best they can be called apparitions. Ghosts!
Siddharth Pai is founder of Siana Capital, a venture fund management company focused on deep science and tech in India.