How to Stride with A.I. in the New Decade

In the coming few weeks of the new year, I will share a series of tried-and-tested principles for successful and impactful application of A.I. within mid-large firms.

____________________________________________________________________________________

We will cover the following topics as primers, for a more advanced series to follow later.

• Foundation, through the Right Motivation

• Risk, and The Shifting Landscape

• Strategy, Frameworks and Platforms

  • A Note on Talent

  • Mitigating Risk

• A.I. Driven Products

  • Strategy for Ideation and Planning

  • Examples

____________________________________________________________________________________

Week 1: Foundation, through the Right Motivation

By the year 2021, no one is unaware of tech driven disruption of industries and companies traditionally considered as stalwarts, and yet succumbing to the inertia of an aircraft carrier; think Yahoo, Blackberry, Sears, Radioshack, et al. Radioshack deserves a special mention because of the nature of its misses. Radioshack missed,

• the tech revolution of a smart(er) phone that could replace many of the devices that it sold, as well as,

• its customers’ changing preferences to source gadgets and parts elsewhere.

To an extent, we are witnessing the same technological revolution continue and accelerate through AI today.

While enterprise agility is first and foremost a function of dedicated and motivated leadership, the path to dynamism traverses a collaborative and positive company culture where employees work for each other, their clients, and don’t have to be divorced from their inherent curiosities. One only needs to carefully observe employee empowerment, success measurement, and rewards, to understand the underpinnings of such a culture at a firm (for more on talent and culture of innovation, refer to this Wharton Study). The ethos of greater social good are also symptomatic through the company’s philanthropy. However, the way a large firm treats its customers can create positive societal waves that may be even greater in their impact. Therefore, ‘customer satisfaction’ from the Social pillar, unlike many other factors within the ESG movement, is both an endogenous (or traditional) as well as an exogenous factor to financial and social success of a firm.

__________________________________________________________________________________________________________________________________________________

Week 2: Risk, and The Shifting Landscape

If AI is no longer an infant, it is the toddler that can outpunch the Muhammad Ali’s (think IBM Deep Blue) of their day. For a specific application of AI, like any upstart endeavor that depends on emerging (some might even say revolutionary) technological paradigm, there is a significant risk of failure along with the probability of a high return. But, for better or worse, this equation also describes the state of our leveraged economy for the foreseeable future. As government mandates, investor and customer attitudes shift in a fully rounded, connected world, we will continue to mature as a knowledge-based economy (even with the devaluation of the dollar) where those who don’t shy away from uncomfortable innovation will continue to take the cake. Returns from AI and innovation are mostly uncorrelated with market returns because of

• the multiple levers of value generation,

• typical VC and private equity investment and exit cycle duration and (il)liquidity, and

• the incentive for innovation during recessions corroborated by the high rates of startup formation during deleveraging periods in the economy.

Success of companies like Tesla indicate a new foundation for aspirational innovation that is provided by the masses, rather than solely by the traditional investing class. While the VC and private equity pockets continue to grow deeper in order to serve as fertilizers for the promising seeds among us, it is the millennial mindset to investing, albeit sometimes misplaced, that reflects the seismic shift in behavioral finance. However, the socially responsible dance of innovation through AI is a tight-rope, one performed well by the likes of Elon Musk. As A.I. determines the winners of the next tech revolution over the coming decade, the political fight for its control is also bound to stay at a fever pitch. As such, AI and data driven innovation must affect a measurable change in at least a subset of the factors underlying the ESG pillars, as well as the bottom line, for an aircraft carrier to dock unscathed in 2030. The surging seas will continue to be unforgiving in the spirit of free & fair markets, and ultimately, human progress.

__________________________________________________________________________________________________________________________________________________

Week 3: Strategy, Frameworks and Platforms

Proper and impactful adoption of AI that does not create additional risk for the firm necessitates attention on the pillars of talent, process, technology, data, and a sound ideation of applications. A strategy that anticipates and survives a vast majority of diverse adversities on the path spanning several quarterly P&L’s is needed. The pillars reinforce a structural change so that A.I. augmented intelligent software can flow through the veins of the entire organization (watch this space for a framework to help formulate such a strategy).

A Note on Talent

We continue to need talent that can work with the nuances of a complex and broad scientific field that is growing at a dizzying pace without creating additional liability. Such experienced talent is stimulated by both, fundamental research as well as applied impact of A.I.. It has both the inclination and the expertise to swim in the deep. As such, firms should carefully facilitate the intellectual curiosity through a combined research, development, and application mindset. Firms that mature their own AI in the new decade have a windfall competitive advantage in terms of operational efficiency and data-driven intelligence.

As A.I. starts on its first spurt of maturation (industry adoption) as a technology now, it will not become an adult in one fell swoop. Much like evolutionary speciation, there will be multiple ebbs and flows, likely over several decades. During the ebbs, it will mutate through basic research to find the right formulations and architectures that are ready for scale (speciation) in the next maturation phase. Leaders who didn’t write-off A.I. as just another IT component are the growth leaders now, and will again be in an enviable position during the next phase.

Process and Tech Backbone

As the cloud computing providers quickly took the lead on de-biasing A.I., which learns from the nature of our social data (and therefore is prone to the same weaknesses as humans), the prolific nature of open source implementations, many of them provided by big tech (see https://www.microsoft.com/en-us/research/theme/fate/ , https://www.microsoft.com/en-us/research/theme/fate/ ), can help solve the foundational challenges of using AI not only legally, but ethically. However, this forms only one component of a risk-mitigating, end-to-end, AI development and deployment platform, which must also include governance, guardrails for budding data scientists, simulation, testing, measurement, automated reporting, proper-use training, operational feedback, and agile iterative design. This reflects our workflow when developing AI prototypes while following engineering best practices guided by Occam’s razor. Such a prototyping framework needs

• shade of the robust enterprise strategy as a prerequisite umbrella,

• a committed partnership with key leadership, and

• AI visionaries with deep expertise.

__________________________________________________________________________________________________________________________________________________

Week 4: A.I. Driven Products

Strategy for Ideation and Planning

As a horizontal, A.I. is set to be as ubiquitous as data in the digital age, with applications in every industry vertical and every function within an organization. Specific goals should be defined to deliver tangible value to the business derived from two broad categories:

• Revenue generation

  by launching new products, creating better direct engagement with customers, providing decision support to the domain experts for better outcomes, and/or

• Cost reduction, through operational efficiency

  After leadership education and ideation sessions have yielded a list of potential use cases, these need to be prioritized based on quantifiable value that they represent. The following steps will help prioritize.

• Calculate the business impact for each use case. Assign a dollar value to either the revenue generation or the cost savings. Think about multiple KPI’s and think long-term (multiple years).

• Assign a probability for each of your estimates above, based on multiple factors such as data & quantitative feasibility, legal constraints, implementation complexity, etc.

• Calculate the cost in achieving these use cases. The sources of the cost are people (hiring and/or training), data, and technology.

Once you select the top X% from prioritized list, factor in other variables, like your strategic initiatives, evolution of your business model, and the inevitable interactions between newly developed/planned products. Some capabilities are more foundational than others, and can be leveraged across multiple other use cases. Lack of such ‘meta’ A.I. formulations is yet another reason why a short-term and organizationally fragmented perspective will create an A.I. backbone where new fractures arise on a daily basis.

Examples

Consider some of the prototypes (on solutions page) from our pipeline as examples of diversity of impact inter- and intra-vertical. Such applications become a reality only through proper strategy, design and execution through a well-defined yet malleable prototyping process and platform.