Chatbots might be making some white collar work obsolete, and this means we will have to innovate — shape up or shape out

The proliferation of chatbots has seemingly occurred right under our very eyes. Late in 2016 Facebook announced there were 34,000 chatbots on its messenger platform. Chatbots are designed to make our everyday lives easier: From booking flights, ordering a taxi, buying a pizza and recommending credit cards, chatbots are built to deliver their services efficiently and accurately using pseudo-human responses.

But how about the use of chatbots in the business world more widely? How can they make me better at my job?

An everyday example

Imagine a scenario where a product manager at a telco wants to build a business case for a price change, or to launch a new concept. Typically they would weave together this business case based on various inputs: current and past business performance, competitive environment, econometric forecasting and customer insight. You can envision the amount of work required to collate this and the level of (human) expertise needed to synthesise the information.

Now this scenario re-imagined would mean a marketer or product manager simply needs to brief a bot on their objectives and all the relevant information needed for their decision is developed within minutes.

The intelligent bots

We are at the stage where some of these tasks can be achieved by AI, even if they are performed at the more elementary level for now. There are already some bots that have the capability to write news, analyse data and provide a simple report, and conduct preference testing for market research purposes. There are even some in development that will be able to review web and print sources, pull out only relevant articles, summarise them and then produce a daily report.

Also read: Chatbots will change the way we do search

And there are bots like X.ai, Clara and Evie who act as personal assistants, work more efficiently and accurately (and never sleep!) but come at a fraction of the cost. All this is quite alarming for the project managers, analysts, external consultants and market researchers who usually form this network of advisors.

Ask Sora

Take “Ask Sora”, a chatbot that sits on the Facebook messenger platform collecting customer insights. The bot performs other activities that engage the user but I want to focus on its ability to capture preferences.

Its primary role is to help people or businesses make decisions through online polling. In response to questions like “Grab Offer or Uber Offer”, respondents vote on what they prefer. This simple, intuitive tool can help businesses and marketers who are often looking to test new ads, concepts, and price points quickly with consumers. Long turnaround times from traditional research firms can be shaved down to a matter of hours or minutes.

A marketer can simply download the application on their desktop, input various concepts or price points, and go to a specific audience for their votes. Within an hour a response will come back telling the marketer which concept is the winning one.

The next stage would be for the programme to sort the responses by age, gender, education and income levels, for example, and also to delve into the “why” behind the responses.

I recently voted on a pair of promotions which definitely had a commercial bend to them. The promotions were offered by Uber and a Grab- mirroring a real life scenario. Below shows the output from the chatbot which reveals that the Grab promotion was more popular!

Wordsmith

Another application, Wordsmith, an artificial writer developed by the North Carolina-based company Automated Insights, can pull out the key elements from a dataset, and generate narratives that resemble a news story written by a journalist. Just by keying in data on income, earnings per share and total revenue, for instance, the programme can convert the data into a full blown article.

Also read: From chatbots to intelligent things: Here are 5 exciting industries startups can focus on this year

The programme can churn out articles faster than any human and is arguably more accurate in terms of focussing on the relevant data points. For example, a human writing data-based reports may miss out on some relationships between key data points as they are prone to subjectivity and other cognitive biases, something which a well-written algorithm is programmed to avoid.

Automated Insights has already gained traction in the market. In 2015 they produced more than 1.5 billion pieces of content, up from 300 million in 2013. And they have several high-profile clients, including USA Today and Yahoo news.

Below is an extract from their website which demonstrates how it converts data into a human sounding story.

However, there is a limitation to this AI. Currently it is not able to provide the “so-what” behind the story, as it is limited in its ability to contextualise and infer, so ultimately what it means for the marketer in terms of the key business decision is unknown. At this stage it can describe what is happening but it cannot steer you in one particular direction or another – this missing piece of the puzzle still requires human judgement and decision making.

A trusted advisor

Ask Sora and Wordsmith are examples of AI that may help to automate some tasks for a business person, but as mentioned they come with some limitations (at least in the short term).

However the notion of machine learning would mean that over time any bot would learn more about the requests and become better at delivering “optimal” solutions. In the Ask Sora case, through a combination of machine learning and improvements from the developers, we would be able to compare results with past iterations, go back to voters for more insight, develop preferred ways of representing the data, and even incorporate external data points. In the case of Wordsmith, the AI could improve on its ability to contextualize and infer and even make some recommendations on the back of the data. In other words, these bots could work up to be a fully fledged consultant.

The issue here is that many of these tasks are currently conducted by white collar workers. As more and more AI pervades the workforce the question arises as to how many of the jobs that require some “smarts” will be displaced. We have already seen the case of BlackRock Inc, the world’s largest money manager, recently announcing that around 13 per cent of its portfolio managers will be laid off as they transform into a business that leans on automated solutions and recommendations based on computer algorithms. As Jerry Kaplan in his book “Humans Need Not Apply” put it so succinctly, “Automation is blind to the colour of your collar”.

—-

Want to be part of the ecosystem?

Register for your Echelon Asia Summit access pass now! Enjoy +10% off Echelon Asia Summit Startup, Investor and Corporate passes just for being our favourite e27 reader.

The views expressed here are of the author’s, and e27 may not necessarily subscribe to them. e27 invites members from Asia’s tech industry and startup community to share their honest opinions and expert knowledge with our readers. If you are interested in sharing your point of view, submit your post here.

Featured Image Copyright: yarruta / 123RF Stock Photo

The post AI might already be stealing our jobs, but chatbots can also help us improve at work through better use of data appeared first on e27.