As we look to create new foundations for building stronger, more resilient economic and social models for the post-COVID future (which, let me assure, you does exist), this new interlock of next-generation technologies has the potential to create next-generation levels of value.
First, let’s examine the four technologies in isolation. Then we will extrapolate their impact as their stack forms.
The idea of a synthetic, digital rendering of a human has been a mainstay of science fiction since time immemorial. Now, that science fiction is becoming science fact. Companies like UneeQ, Soul Machines, Unreal Engine and LG Electronics have created digital people with a veracity and verisimilitude that takes them far across the uncanny valley of prior screen-based robotic representations.
To the casual observer — i.e., unless you’re really, really, really squinting — digital humans are indistinguishable from real people. Able to interact over a screen to handle a service issue or a customer inquiry, these instantly and always available digital presences are good enough to be “good enough” to handle L1 and/or triage interactions in a range of scenarios. This is particularly true in a world of Zoom and other low-to-medium-resolution videoconference-based interfaces.
Machine learning scripting software
Although long over-hyped, the true power of artificial intelligence (AI) is still under-appreciated. The AI-driven language generator GPT-3 is changing that. Anyone who has played with the latest version of OpenAI’s autoregressive language model is literally sockless — their socks having been blown off by its ability to create streams of language that are convincingly human. The famed Turing test — the British mathematician Alan Turing’s suggestion that we will have a “thinking machine” when its responses are indistinguishable from a human’s in a blind test — is now decidedly in the rearview mirror; GPT-3 is an exponential leap forward in taking deep learning into a new chapter of impact and potential.
Exclusively licensed by Microsoft from the Elon Musk-funded organization OpenAI, GPT-3 will soon be woven into many different Microsoft products; in time, one can imagine Word offering auto-suggestions of full sentences, paragraphs and even chapters of text, based on keywords and objectives required by an “author.”
Robotic process automation
RPA software has been around for some time and is forecast by Gartner to be an almost $2 billion market in 2021. Vendors like Blue Prism and UiPath have established sizable beachheads in organizations looking to automate rote, repeatable elements of business processes, such as approving a mortgage or identifying fraudulent insurance claims.
With pressing imperatives to reduce costs and de-risk process execution against exogenous interruptions (such as the COVID-19 pandemic), however, the adoption of RPA (and the next iterations of RPA that will have machine-learning approaches baked into them) is set to increase markedly in the next few years; Gartner projects enterprises will triple their use of RPA in the year ahead.
Although, as with AI, it’s subject to hype and supply-side exaggeration, RPA is a technology (and approach) of great promise — predicated, as it is, on the fact that so much activity within large organizations is potentially automatable. A huge amount of white-collar work undertaken in cubicles (and now at kitchen tables and converted bedrooms) around the world is standardized and “if-this-then-that” in nature; that this type of work hasn’t been automated so far is typically a failure of imagination and middle management’s ability to see beyond short-term business-as-usual.
Quantum computing — the most nascent of the four technologies of the new stack — has the potential to make the biggest impact. In late 2019, Alphabet (aka, Google) published peer-reviewed research that demonstrated a quantum computer had solved a calculation in 200 seconds that would have taken a classic supercomputer 10,000 years. From 10,000 years to 200 seconds. Let that sink in.
On this current trajectory, a huge range of new possibilities come into focus: modeling new vaccines and medicines, as Boehringer Ingelheim is doing, predicting (with complete accuracy) economic outcomes, simulating new materials, creating new communications standards, manipulating genetic material (with a turbo-charged CRISPR) and developing fully autonomous vehicles, as some examples.
Although early in its commercial viability, quantum computing will, over time, become a more mainstream commercial offering. At that point, a new arms race of processing power will have started, making our current supercomputers look no more capable than the ENIAC and Apple 1 of old.
Power in numbers
In isolation, each of these four components of the DIGITALL stack has an important and sizable future. In tandem, they are set to be transformative.
Imagine a digital human, instantaneously customizable by ethnicity, age, aesthetic, language, gender — powered and voiced by scripting software (leveraging natural language processing to understand the spoken and written word) that can craft real-time responses and prompts — powered by RPA software that understands a process “journey” and guides a user (a customer, an employee, etc.) through a complex series of steps in an elegant and calming way — powered by a quantum computing engine operating at beyond-human-synapse speed.
This DIGITALL stack is a logical endpoint of technology trends that have been gestating for, in some cases, decades but which, due to their inherently overwhelming qualities, have been easy to overlook.
The combinatorial and transformative power of the DIGITALL stack can be illustrated by recalling any online or phone-based customer service interaction you’ve recently had. Did you come away from it satisfied, happy, your mood enhanced, your burden lifted? Of course not — you came away frustrated, irritated, annoyed at how long it took, how many times you had to repeat yourself, how you had to provide the same information again and again, how the agent didn’t seem to understand you, seemed to be following (badly) a script, sounded almost robotic.
Stacked with answers
Recently, I had that type of experience, spending hours trying to fix an issue with my broadband provider. Without naming names, and with all due respect to the person on the other end of the call who did their best to help me, the experience was completely sub-optimal/frustrating/painful. The current bar of human performance is not going to be that hard to leap.
Estimates vary as to how much money businesses lose through bad online and phone-based customer service, but the consensus is that it’s in the billions of dollars each year. Fixing this longstanding and ongoing problem through traditional means (better training, better HR policies, etc.) has clearly failed — a new approach is badly needed. The DIGITALL stack provides such an approach.
Digital humans, continuously updated with customer and market insight and intelligence gleaned from millions of other transactions undertaken by other digital humans all active within the same operating system, powered by quantum technology, could completely upend a business process that is prime for reinvention. With COVID reminding us that necessity is the mother of digital reinvention, this reinvention is more pressing and called for than ever.
Prepping now for the DIGITALL of tomorrow
Of course, the application of the DIGITALL stack will not be limited solely to customer service — it is potentially deployable in any area of human-to-human non-in-person interaction: seeing a doctor, consulting with a tax accountant, watching the news, receiving your annual work progress evaluation.
With an appreciable dose of human oversight and governance, machines will begin to write as well as humans and without any of the costs and complications of humans … and talk better than humans and look better than humans and work all of the time (and not be worried about COVID) … and get better and cheaper every day, and not need to be housed in expensive real estate in expensive cities. When more and more of our life and work is online so that the human advantage of presence and reality has diminished — it’s then that the promise and benefits of a DIGITALL stack will begin to become more and more apparent. That point is almost here.
A new stack means a new set of roles and rules — rules that will be key to your organization’s future in the years ahead, based on dramatically lower costs (the largest cost on most balance sheets is employee salaries), dramatically higher net promoter scores (contact your own organization’s customer service desk under a pseudonym; afterwards, ask yourself (honestly) would you recommend that organization to your significant other?) and very dramatically faster processing times (“the loan you applied for a moment ago, Madam, has just been approved”).
The new DIGITALL stack is at an early stage of its S-curve lifecycle. Outsized rewards will come to those organizations that understand this new paradigm and act to operationalize it, at scale, judiciously. With the imperative of not letting a crisis go to waste, and with the objective of deploying technologies that will withstand the next black-swan event just around the corner — recalling the famous maxim, “life is just one damn thing after another” — the DIGITALL stack will prove important after the virus and far beyond.
This article was written by Ben Pring, Managing Director of the Cognizant Center for the Future of Work.
To learn more read the latest Cognizanti Journal, "The resiliency reset," or contact us.