Are chatbots in the workplace the entry point to cognitive personal assistants?

HR chatbot

Like robots, artificial intelligence and machine learning, one topic getting a lot of attention at the moment is chatbots.

Much of the news and commentary about these relatively simple computer programs have been on the negative side – Microsoft’s Tay, for instance, and the blanket coverage of the conversational meltdown that happened within days of the chatbot’s public trial in March. Not a good pointer to the utility and value of such algorithms! (I noticed, btw, that the Tay URL currently gives a DNS error rather than displays the site itself.)

My take on this is a simple one – what we’re seeing now is just the tip of an iceberg where literally anyone can create a chatbot and make it openly available to meet a need where the tool and how it’s offered – text messaging typically via smartphones and also via computers with a computer algorithm that uses natural language understanding capabilities – are a large part of their simplicity and, hence, utility and appeal. There are chatbot directories where you can search for and find chatbots.

Among the myriad commentaries and opinions on this and related topics, I’ve seen a few just in the past week that focus on areas that should be of distinct interest to strategic business planners, HR practitioners and communicators (among others: the C-Suite, for instance).

One example that especially caught my attention is that of online retailer Overstock and its chatbot known as Mila, developed by customer experience platform developer Aspect (the company originally known as Voxeo). What’s interesting is who the users of this chatbot are – Overstock employees, not customers as you might assume.

Bloomberg reports:

Starting this July, when customer service employees are too sick to come to work, they’ll tell Mila they’re not feeling well using an app on their phones or computers. “I’m sorry to hear that,” Mila will respond. After a short exchange about logistics, Mila will send a message to the appropriate manager, who will adjust the employee’s schedule.

[…] Mila will replace a more cumbersome call-in hotline for ill employees. Currently, when any of the 450 employees in Overstock’s Salt Lake City call center feel under the weather, they pick up a phone and leave a message. Someone has to check those messages and tell a manager, who replaces the worker on the schedule. Chat’s way is easier: Now the company can fill schedules and replace workers faster, which ultimately saves money. “Those seconds for each employee turn into dollars,” said Stormy Simon, president of Overstock.

In addition to chatting in sick, employees can use Mila to schedule time off, check their schedules, and do a variety of other tasks that used to require making a phone call, sending an e-mail, or (gasp) talking to a fellow human.

The image above shows an example of such a “conversation” between an employee and the Mila chatbot. It sounds all so easy, doesn’t it? Automate a routine HR procedure like employees calling in sick in a way that’s easy, appealing to the user of the new tool, that improves a business process through automation, and saves money.

You can project this chatbot idea out further to take on other routine internal tasks and processes common to many businesses.

Imagine its use in employee communication, too, to offer services to employees in new and interesting ways via their mobile devices, in tune with what people increasingly have become accustomed to and expect, and forms part of their routine in how they receive and interact with certain types of information.

Imagine, for instance, being able to ask your chatbot for a summary of your health benefits. Or maybe links to comparative information on a range of cars that you’re thinking about for your next company car. Perhaps some configuration details on how to connect to a printer in a branch office you’re visiting. These are the types of tasks that lend themselves to this use of such technology to replace cumbersome, relatively slow and labour-intensive methods like telephone, manual intranet searches and email.

If you use Twitter, you may already have experienced a chatbot. Slack, too. And Facebook who announced chatbots for Messenger earlier this month.

I think text-messaging chatbots will become a feature of the organizational landscape in many companies by the end of this year as awareness grows of the value of such tools internally and the relative ease and low cost in how they can be created, and as use cases emerge from which others can learn.

I also think that the hype build-up will be huge, setting the scene for equally huge disappointment when everyone’s expectations of what they think chatbots can do meets the reality of what they can actually do.

Yet this is just the beginning of where the real value will come from automating processes that provide significant benefits to employees and the organization that go way beyond text-messaging chatbots.

The Next Level

What I’d like to see is the means by which my whole daily routine is managed for me by my assistant – my cognitive personal assistant, my CPA. It gets to know my preferences, my likes and dislikes, my behaviours so that it is able to manage that routine.

I don’t just mean my asking it things. I expect to do that, of course, but I also expect the CPA to anticipate my needs and questions and already propose solutions for my approval. Over time, I’d allow it to actually make decisions.

So all this could translate into routine things like these, just a handful of examples:

  • My CPA will manage my calendar, accepting and rejecting meeting requests, and organize meetings and invite participants.
  • It will filter my email for me using its knowledge about me and my preferences and – more importantly – knowing about those who send me email and their relative significance to me depending on factors like who they are and where they are in my organization and in others.
  • It will take care of everything regarding my business travel: airport parking, flights, taxis, accommodation, all of it.
  • It will remind me of things I need to give my attention to, ranging from meeting reminders to my to-do list.
  • It will be my valued research assistant.
  • It will recommend things to me based on its knowledge of me and my interests, all relative to my daily routines and what I’m working on.

And one more:

  • My CPA will chat with chatbots for me.

My interactions with my CPA will embrace all forms of communication – written, audible and visual. Whatever is the appropriate method.

I’ve spoken about much of this in various ways at a number of speaking events this year, public and private. A great deal of my current thinking is richly informed by what I learn from seeing and talking about at IBM (my employer), particularly where IBM Watson and collaborative platforms are concerned, and what’s in the thinking of clients and where they see this going.

I am sure that all of what I think of as a CPA won’t happen all at once. If you think about it, even the short-list of examples I mentioned above represents quite a leap in terms of mindset shifts required in organizations, never mind the computing challenges to create the frameworks for it all to happen, seamlessly and reliably.

But from little acorns do mighty oak trees grow.

Additional reading: