Artificial Intelligence

What happens when you work with a robot colleague?

The study: An AI system was fed medical records of all 1.4 million visits by 567,498 patients to a medical center in Guangzhou in China. All of who were aged under 18. A team distilled this information into keywords linked to different diagnoses, then fed these into the system to help it detect one of 55 diseases.

The system managed to diagnose conditions as varied as common ailments like influenza and hand-foot-mouth disease to life-threatening conditions like meningitis with 90% to 97% accuracy. Its accuracy was compared to 20 pediatricians. It managed to outperform the junior ones but more senior doctors had a higher success rate. The findings are described in a paper in Nature Medicine. (Reference – Evaluation and accurate diagnoses of pediatric diseases using artificial intelligence by Huiying Ling and Brian Y.Tsui).

In the above example, one can be fairly confident that the algorithm would soon perform at levels of the senior doctors as well, if not better. Increasingly offices all over the world will be filled with computers and bots who will be super-efficient when compared to the average worker. Then as a worker how do you work with someone who is much superior to you in certain aspects?

Let us take an analogy on how technology has started impacting the workforce – While it is true that telephones have existed for more than 100 years now, mobile phones have completely redesigned our lives over the last few years. The influence of the bots on the shop floor is something similar and Bot-Human relationship is something for the management to look very closely at.

Before we get to that, let us see what Dilbert would say about this future world!

Dilbert Cartoon via

This article aims to look at how would the evolution of the office space could look like in a world where computers/bots are our colleagues, subordinates and maybe even our managers!

Some thoughts are given below:-

Turbo charging business decisions

If we go back a little in time to period when Microsoft Excel was not very popular, there used to big calculators used for computing numbers at the end of each day in various functions. Once we had Microsoft Excel, we were using that time into making graphs and diagrams based on the numbers thrown out by the spreadsheet. Now bots have effectively reduced the “Average Handling Time” of many processes and made humans spend less time on Data Collection and more frequent decision-making cycles.

Event-based processes

Most of the processes so far have been designed in a time-bound way as humans want a minimum number of transactions to have been completed before they report out status change. For e.g. Weekly Reporting, Month End Accounts Close etc. However, all this could change as now Bots can work 24*7 and be available to report out the status instantly. So it is highly likely that the management would want to move to an event-based process rather than a time-bound For e.g. Automation would allow the organization to estimate expenses incurred at any given point of time and compare it against the budget rather than waiting for the month end to know the variance. We have already seen this happen to our bank statements where real-time status is available for us to check.

Managers will become work flow designers

There is a term called “Grunt Work” – Work which no one wants to do as it is mind-numbing and highly repetitive. These are ideally delegated to the bot. The parts of the job which involve “Judgement / Expertise” should ideally be the human focus. As in every successful relationship, bots and humans should complement in the office playing to each other’s strengths

Humans to lay down the highways

Bots are like Cars and Process Standardization efforts carried out by the Humans lead to the smooth roads needed for being able to drive the car and enjoy the ride. This is one of the key success factors for successful bot implementations and leads to better yields in terms of cost benefits.

Managers have to go beyond their silos

Managers should take up an end to end view while integrating the work between their team members and bots. Bots are generally implemented in a particular team and are highly dependent on upstream inputs / downstream process efficiencies for effective performance. A bot needs human support to provide the necessary inputs and environment to function effectively. E.g. If a bot has consistently poor data quality issues in the invoices being processed leading to poor utilization of the bot, the humans in the team can reach out to the upstream functions ( e.g. procurement) and try to set up Cross Functional teams to fix the issue.

Employees should move towards becoming techno functional resources

Organizations have to invest in building Techno-Functional resources. The role of a bot-operator is technical but there needs to be a deep functional understanding of the domain as well. This will help humans to improve the bots over a period of time and continue to achieve better business outcomes.

Changing role of experienced resources

Each team would have a senior member who has “been there and done that”. In the new workplace, the experienced resources would be expected to share wisdom/strategy rather than their experience, as the bots would have digital imprints of previous instances of the same transaction. Also a few years ago the junior staff would have found mentors/guidance from the senior members in the team. With increasingly smaller teams due to automation, the organization might have to invest in mentorship programs to mentor the junior staff to grow in their careers.

Focus on what can go wrong

As every movie on technology has warned us already, technology gone wrong can have damaging consequences. It is true that bots have to be managed carefully. The damage due to a rogue bot can be many times more than that of a rogue employee. It is the humans who would be expected to take charge of the controls, checks, and balances involved in running the bot.

Turn the clock back once in a while

Over-dependence on Bots is not desirable from a BCP perspective. The humans in the team would have to run the processes manually to ensure that there is an Option B, in case there are bot failures. It will be a best practice to have multiple resources to be trained on a bot so that there could be multiple team members who could jump in when the bot fails / during BCP.

And the best bot of the year awards goes to…

As Peter Drucker once said, “what gets measured, gets done”. This is equally true of humans and for bots as well. Visual dashboards in the shop floor have to reflect the performance of the humans and the bots. Eventually, the bot performance will permeate into team meetings similar to human performance and issue resolution. However, it is important to measure human contribution and bot performance with very different yardsticks.

To summarize, an evolving work floor comprised of humans and bots is going to be one of the most interesting organizational dynamics problems for all of us to observe and participate in. Progressive HR Teams have to start building policies around how to manage interaction with Bots. Maybe they’ll use cognitive bots to do that too!

Leave a Comment

Your email address will not be published.

You may also like

Pin It on Pinterest