The robots are coming, but the humans aren't leaving Fernando Aguirre de Oliveira Júnior Partner Services, Outsourcing & Automation Advisory May, 2017
Call it what you want, digital labor is no longer a consideration But now a mandate. It is no longer about if, but about where, how, and how fast. 2
Cognitive Technology and Digital Labor Cognitive systems are an application of interpreting and learning systems and redefine the relationship between human and machine $152B+ The expected market size for digital labor by 2020* Cooperative 2017 KPMG ( KPMG LLP, a International ), Delaware limited a Swiss liability entity. partnership All rights and reserved. the U.S. member firm of the KPMG * network Bank of of independent America member Merrill firms Lynch, affiliated with November KPMG International 2015 ROI between 600 800%* 45% of activities individuals currently perform in the workplace can be automated using existing technologies* * London School of Economics, The IT Function and Robotic Process Automation, October 2015 * McKinsey & Company, Four Fundamentals of Workplace Automation, November 2015 3
How do we address the potential impact to human labor? How can we use Digital Labor to our advantage? 4
Part of a new Industrial Revolution? 1800s Steam, water, mechanical production equipment 1900s Division of labor, electricity, mass production 1960s The Information Age - Electronics, IT, Mass Communication, Internet Today AI, nanotech, biotech, cyber-physical systems, digital labor The 4 th industrial revolution (cyber-physical systems) is beginning and its impact is profound according to the World Economic Forum founder, Klaus Schwab 5
Cognitive Technology + Digital Work = Digital Labor Cognitive Technology (aka Artificial Intelligence): simulate the way humans perceive, learn, reason and respond Digital Work is the human task of organizing data and applying human context Digital Labor is the valorization of digital work to automate activities and tasks that previously required human labor* *Influenced by works of Christian Fuchs and Sebastian Sevignani of the University of Westminster 6
The scope of Digital Labor is broad Spectrum of labor automation to augmentation The tools used range from traditional automation software to new cognitive platforms which make automation tools intelligent and help augment and leverage human knowledge 7
Frictionless access Exponential technology Human knowledge shared freely Global demographic shifts to technology by more than 2 billion people (mobile, cloud) improvement growing at more meaningful baselines on the internet is giving context and meaning to digital content reduction in working age population and need for talent Why now? 8
Technology underpinnings of Digital Labor? Machine Learning A mature field of computer science where algorithms learn by statistical methods associated with datasets and the associated meta-data. The current focus is on learning with exponentially larger data sets as presented by IoT and image and video data. Digital Labor Cloud Computing The ability to access low cost, high performance computing from any device has dramatically transformed how data is created, processed and consumed and largely enabled rapid advancement of AI. Deep Learning Advancements in neural networks that are trained by humans to perceive abstract concepts in text, images, sounds, and video has dramatically advanced practical applications in Natural Language Processing, Computer Vision, and machine reasoning. Big Data and Social Media The large sets of data supported by human context through social media channels are driving unprecedented knowledge and insight in a digitally consumable way which fuels and accelerates AI. 9
Cost Efficiency Estimates suggest that a software robot is approximately 1/3 of the cost of an offshore FTE. Digital labor savings are estimated to be between three and ten times the cost of implementing the automation Productivity/ Performance Software robots work 24/7, and 365 days a year; do not take vacations; and perform tasks at digital speeds Consistency/ Predictability Expected reduction in mistakes, accidents, regulatory violations and fraud Business implications of Digital Labor Quality/ Reliability Software do what you tell them to do when properly configured they do not make mistakes and thereby eliminate human error Employee Satisfaction & Innovation Eliminating mundane and repetitive tasks frees up human talent to innovate and create Scalability Software robots scale instantaneously at digital speeds to respond to fluctuating workloads. There is also no overtime, no hiring challenges and no 10 training.
Classes of automation Automation of entry-level, transactional, rule based & repeatable processes Characteristics: Working with structured data and within well defined parameters, virtual robots can complete tasks autonomously These tools sit at the presentation layer and do not infiltrate the IT system Solutions are easily designed, quickly tested, and implemented with a relatively low investment or expenditure Human factor replaced. Example: A US-based online bank has used RPA to automate tier 1 inquiries (i.e. address changes) Processing of unstructured data to support elements of self learning Characteristics: Enables the capturing of process knowledge, and applies this knowledge to instruct how the process should run Based on evidence, defined process outcomes are generated. These consistently carry a high probability of the desired output Speeds up human analysis to drive the right decision. Example: An energy company utilized AI and advanced semantic reasoning to deploy a virtual service desk agent to rapidly understand questions, provide customers with answers and escalate to humans if needed Automation driven by self learning and adaptive technologies Characteristics: Can be used for sophisticated cognitive hypothesis generation/advanced predictive analytics Such platforms are costly to develop and implement, and generally require a long lead time currently very few players Reduces human error, but does not take humans out of the equation Example: IBM Watson s natural language processing, machine learning, pattern recognition and problematic reasoning algorithms are aiding skilled employees with complex decisions 11
Example: Class 3, KPMG teaching IBM Watson to support audits 12
Platform Business Models? What s the security risk? Impact to Labor and Talent? Ethical and Social Issues? What are the strategic questions? 13
Platform Operating Models Fundamental changes to operating models from human centric to technology centric Cloud platforms offer the technical means to cognify and automate an activity Engagement models with suppliers and customers change to digital channels Businesses must develop a long term roadmap with nearterm- mid term and long term priorities Example: People centric outsourcing to Cloud Based platform model 14
Cyber security Going digital means moving more into the cloud The power of machine learning and digital labor can be used in malevolent ways Cyber security must evolve in scope and consideration of the leveraged data and computing capacity available to criminals. 15
People and talent Experts are required to train and oversee design, content curation, data analytics and technology development and improvement on the platform i.e. Digital Work Talent becomes more critical as a differentiator as many of the routine activities are automated at a low cost and skill, innovation and agility becomes the competitive advantage Technological unemployment may occur in lower skilled areas but demographic shifts are putting pressure on labor supply and demand Cooperative 2017 KPMG ( KPMG LLP, a International ), Delaware limited a Swiss liability entity. partnership All rights and reserved. the U.S. member firm of the KPMG network of independent member firms affiliated with KPMG International 16
Demographic changes Technological unemployment has been a concern since the 1700 s Is this time different? There are clear areas of concern in routine cognitive jobs But the global population is aging and departing the workforce which is causing over-leverage of pensions and benefits regimes In developing countries, the source of recent growth is subject to automation 17
Demographic changes Brazil will continue to buck the trend for perhaps another decade 18
How to deal with job disruption and rising inequality? Ethical Questions Is Universal Basic Income (UBI) the answer? Who is responsible for mistakes that machines make? How to influence and deal with public policy and regulations toward AI? 19
Innovation Software Development and Data Sciences Where will the new jobs be? Design, Curation, Training and Operations Cyber Security 20
Many areas are impacted here are a few Transportation Autonomous vehicles Air traffic control Cargo and logistics Drones and small package delivery Business operations Accounting and Audit Risk Tax Transaction processing Sales and Customer Care Healthcare Drug discovery Diagnoses and treatment Insurance coding and processing Wellness Patient monitoring and care Finance Investment management Securities trading Banking operations Fraud detection and prevention 21
Human value will remain unparalleled 23
The next 3 years more disruptive than the last 50? 24
25
Thank you 26
kpmg.com/socialmedia The information contained herein is of a general nature and is not intended to address the circumstances of any particular individual or entity. Although we endeavor to provide accurate and timely information, there can be no guarantee that such information is accurate as of the date it is received or that it will continue to be accurate in the future. No one should act on such information without appropriate professional advice after a thorough examination of the particular situation. 2017 KPMG LLP, a Delaware limited liability partnership and the U.S. member firm of the KPMG network of independent member firms affiliated with KPMG International NDPPS 537977 The KPMG name and logo are registered trademarks or trademarks of KPMG International.