Redesigning work in an era of cognitive technologies
Since the first caveman picked up a sharp rock, technology has been changing the way we work and the jobs we do. What effect will the rise of artificial intelligence and cognitive technologies have on jobs of the future? Host Tanya Ott talks to David Schatsky about the kinds of work computers can do now, what they'll be able to do soon, and how it may affect your business.
The fears about the impact on workers are justified. It’s important more than ever for people to continually refresh their skills, and to make sure they are as flexible as they can be and that they learn how to take advantage of these technologies in their work so that they can be empowered instead of replaced.
Want to learn more about cognitive technologies?
Tanya Ott: This is The Press Room, Deloitte University Press’ podcast on the issues and ideas that matter to your business today. I’m Tanya Ott, and I’ll be talking to thought leaders and change makers about trends in business analytics, CFO strategy, risk and security, social impact, and more. Today – it’s emerging technologies.
You’ve had this experience, right? You’re looking to buy something, but first you want to do a little research. You read some reviews online. Visit a couple of store websites. Maybe you make a purchase, maybe you don’t. But the next thing you know you’re not the hunter – you’re the hunted. Everywhere you go online, that product – or ones like it – are popping up in search results, on the sidebar of your favorite social media sites.
That, my friends, is cognitive technology at work. This is too….
Automated customer service system: Are you still there? If you’d like to continue, say yes, or press 1.”
Tanya Ott: And so is this… “Siri – what’s the phone number for David Schatsky?”
Siri: I don’t see David Schatsky in your contacts. Perhaps you meant one of these 15?”
Tanya Ott: David Schatsky spends a lot of time thinking about cognitive technologies. He analyzes emerging technology and business trends in Deloitte LLP’s US Innovation group, and he wrote the article “redesigning work in an era of cognitive technologies,” which will be published in the Deloitte Review in July.
David Schatsky: “In the old days it was thought that data was something that people created deliberately. Transactions that were recorded, for instance, or letters that were written and saved. But now it’s been understood for 10 or 15 years or more that there’s a lot of data that’s created incidentally – what people call digital bread crumbs or digital exhaust… tracks and traces of activities that people are conducting all the time online. And with the ability to analyze massive quantities of this kind of data and then extract meaningful insights from it, that whole definition of where information is, what it is and how you get value from it has expanded.”
Tanya Ott: Can you give us an example of those digital breadcrumbs?
David Schatsky: “A user of the internet who’s visiting multiple different sites and then purchases a product online. That may seem like a random set of behaviors that culminated in the purchase of a book online, but if you track that kind of behavior across millions of people you might find certain patterns there that suggest that there’s meaning. For instance, people who read a story about some topic, then read another article about that topic – may be they’re 10% more likely to a book on that topic. That may have been a pattern that wasn’t visible without the large amount of data tracking people’s behavior that we have today.”
Tanya Ott: Things are moving fast in the world of cognitive technology. Schatsky and his team of researchers analyzed 100 different applications and found they basically fall into three categories. The first is “product applications”.
David Schatsky: “A product application is where a company imbeds cognitive technologies such as natural language processing or computer vision or what have you in a product to provide an intelligent behavior or a natural interface such as being able to communicate with a product via speech or visually. Examples are things like robotic toys, intelligent thermostats that learn from experience what temperature you want your house at to things like autonomous vehicles that have robotic ability, computer vision, and planning capabilities built in so they can operate themselves.”
Tanya Ott: The second bucket is what he calls a “process application.”
David Schatsky: “Process applications use cognitive technologies to enhance, to scale up or to automate business processes. We’ve been using information technology and before that even industrial technology to automate work in organizations but now cognitive technologies enable us to automate and scale up different kinds of work. So automating data entry for example with software that can automatically recognize handwriting or automating planning and scheduling, automating customer service with speech recognition. Each of these corresponds to a business process an organization might undertake.
Tanya Ott: And finally – the “insight application” … he says this one is really important. Machine learning can reveal patterns in large sets of data – patterns that can make predictions and guide more effective actions.
David Schatsky: “This is a rapidly growing area. But one that we cited in the research is an example of Intel, that has started to use machine learning to guide work by its sales force. So by analyzing the patterns, buying behaviors of its customers, a machine learning system can direct the sales people on which customers to call and what offers to make them. And so, taking out some of the guesswork that sales person faces in how to prioritize their work. And Intel has found that the sales people who’ve adopted this approach have dramatically improved their productivity and increased revenues along the way. “
Tanya Ott: These insight applications can be so powerful that recently a Japanese venture capital firm named an artificial intelligence member — basically a computer – to its boards of directors.
David Schatsky: “It makes for awkward company picnics I think! What they’re trying to say is that’s we’re not going to make important decisions without doing the proper analysis of the kind that our AI assistant is able to do.”
Tanya Ott: So, how does a company decide whether to automate a process? Schatsky says one of the first things it should consider is the Three Vs.
David Schatsky: “The 3 Vs stand for “Viable”, “valuable” and “vital.””
Tanya Ott: Question #1: is the automation technology you’re considering Viable?
David Schatsky: “You have to understand whether it’s possible to apply a technology to a problem. These are ones where we have tasks that computers are now good at: recognizing images, recognizing handwriting, understanding speech in certain types of domains. Planning and optimizing those kinds of things. You have applications like this that might be viable.”
Tanya Ott: Question #2: Is it valuable?
David Schatsky: “Can you do something better in a way that’s valuable to a customer: either faster or better quality or more consistently or at a higher scale. Or can you do it in a way that can cut costs b/c perhaps you have labor whose talents are not fully utilized. Most people spend a percentage of their working life doing work that they’re overqualified to do. Even a person with sophisticated training spends time reading documents and extracting key information from those documents. That’s a task a computer can do. So when you have tasks like that where skilled workers don‘t have the opportunity to fully employ their skills b/c some of the work they’re doing is not particularly skilled work, that’s an opportunity to apply cognitive technologies and automate part of what they do. And free them up to do higher value things. “
Tanya Ott: And Question #3: Is the process vital?
David Schatsky: “In a growing number of areas, companies won’t really have a choice. They’ll have to employ cognitive technologies b/c cognitive technologies will reflect really the standard way of doing business. So one example is fraud screening in payments industries. Nowadays high scale payments processing companies have to use machine learning techniques for fraud detection because that defines the highest quality fraud detection techniques and if they don’t do it, they’re not performing at industry standards.
Tanya Ott: If your company isn’t performing at industry standards, it’s likely losing money – plain and simple. Schatsky says companies have a choice between a cost strategy and a value strategy when deciding how to deploy technology.
David Schatsky: “A cost strategy uses technology to reduce costs, especially by reducing labor costs. If you have a technology that could possibly totally replace a worker with a cognitive computing system, you have a choice of doing that saving the cost of that worker’s salary, but the upside is really limited to the cost savings that you could achieve. The value strategy might be assigning those workers to new roles, or expanding their roles, or helping those workers provide superior performance with those technologies. “
Tanya Ott: Schatsky says not everything that can be automated should be automated.
David Schatsky: “It needs to be undertaken with a view toward the workers who are affected and how you can make choices that will get the most value from them.”
Tanya Ott: Schatsky’s team identified a spectrum of choices starting with replacing some human workers with machines.
David Schatsky: “Examples are automated teller machines or the interactive voice response that you get when you call a company and you only are talking to with a computer.”
Tanya Ott: The next step along the spectrum is to “automize” the work. It’s a mash-up of “automate” and “atomize” – or break up into pieces. Automate as much as you can and whatever’s left – a worker does.
And then there’s the “relieve” approach – where a company identifies the parts of a job a person doesn’t like to do…
David Schatsky: “They’re dull, dirty, of their dangerous. Those kinds of tasks will get automated and you’ll leave the stuff where people can really shine, for people to do.”
Tanya Ott: You can look at this one a couple different ways…. based on that “cost versus value strategy” we talked about earlier.
David Schatsky: “We see it in cases where firms that are doing translation of languages will have a computer do much of the translation automatically and then have the translator just clean up or edit the results, which is something that they hate. So that’s a cost strategy. But a value strategy might be to create new products based on your ability to automate things. So in the translation example, maybe the customers in the past either said in the past either said we can’t afford to translate because doing high quality translation is too expensive or they can. Now, with automation, you can offer different price points. You can partially automate or fully automate. You can mix and match the amount of technology and the amount got labor that you use to address new markets with different price points. “
Tanya Ott: Schatsky says the final automation approach is Empower.
David Schatsky: “And that means automating things that weren’t even being done before by people to make people more effective. So examples are things like IBM’s Oncology Advisor that can digest 100s of 1000s of medical papers to help physicians make more informed choices as they’re prescribing treatments for cancers. No doctor can keep up fully with the medical literature, so this is automating something that wasn’t done before and it empowers physicians to do better than they were doing before.”
Tanya Ott: Schatsky concedes – there are some industries where cognitive technologies are more challenging… mostly because we mere humans haven’t quite caught up.
David Schatsky: “ The legal and regulatory framework might not be fully developed so – something that people talk about a lot is autonomous vehicles and even if you have a car that’s able to drive from one part of the city to another, who’s responsible if there’s an accident? Is it the person that bought the car? Is it the person who programmed the car? Is it the person sitting in the car? There are questions like this that will stand in the way for a while for the adoption of these technologies until the legal and regulatory framework gets sorted out.”
Tanya Ott: There are concerns that the pace of change we’re seeing right now could exceed the ability of society to adapt. Could we end up the players in one of those sci-fi movies where robots completely replace humans? Should workers be worried? Schatsky says “yes” … and “no.”
David Schatsky: “There’s no question that technology changes the nature of work. That it eliminates certain kinds of jobs, and those jobs can be automated. There’s no question that historically technology has created new types of jobs and has created wealth along the way. So at some level what we’re seeing now is potentially the continuation of a pattern that we’ve seen ever since the first technology was created. Whether a lever, a fulcrum or a shovel, what have you. So as we’re starting to see jobs that never could be automated before – jobs that require human perception, human communications, cognitive kinds of skills. As we see those start to get automatable and automated there’s a question about what the people who are doing those kinds of work will be able to do. Nobody really knows what shape the economy will be. What kind of jobs and new industries will be created in the future? But it’s certainly a valid question to ask. From the point of view of workers it’s more important than ever for people to continually refresh their skills, and to make sure they are as flexible as they can be and that they learn how to take advantage of these technologies in the performance or their work so that they can be empowered instead of replaced.
Tanya Ott: We want you to be empowered. That’s what the Press Room is all about. So don’t miss a single episode of this podcast. Subscribe on iTunes or via your favorite podcast app. And don’t forget to leave us a rating and comments. We want to know what you think about cognitive technologies and everything we tackle here in the Pressroom. Send us an email at email@example.com or tweet us @DU_Press. You can find out more about the discussed today at DUPress.com. Thanks for listening, I’m Tanya Ott.
This podcast is provided by Deloitte LLP and is subsidiaries and is intended to provide general information only. This podcast is not intended to constitute professional advice or services of any kind. For additional information about Deloitte LLP and its subsidiaries, please go to Deloitte.com/about.