AI’s coming of age, in the middle of a pandemic

We are entering a new chapter in Artificial Intelligence (AI) adoption in the enterprise. Capabilities are advancing rapidly, case studies emerging every day and research accelerating quickly. Frankly the fight against the Novel Coronavirus and the new normal has changed our dystopian perspective on AI to one that is of a human-machine partnership to combat the invisible enemy. The findings of the latest McKinsey Global Survey on the subject show a nearly 25 percent year-over-year increase in the use of AI in standard business processes, with a sizable jump from the past year in companies using AI across multiple areas of their business. Based on my own conversations with leaders, partners and practitioners in the past year, AI is now a fundamental element of the next generation business fabric. Here are some of the observations, what to expect and how to operationalize AI in your enterprises going forward. This is a two part series, the first one for the business leader in you to understand the macro-movements in this space, and second one, will explore a bit more on the nuanced side of the technology and the operationalization of enterprise AI.

It’s time to double down on your AI investment: It is true that not everyone has adopted AI technologies yet — there are still barriers, and many are working to scale the benefits. That said, AI is moving out of the “early adopter” phase in the market and into the “early majority.” In fact, IDC forecasts that spending on AI technologies will grow to US$97.9 billion in 2023 — more than two and a half times the spending level of 2019. That said, the early adopters are now providing a playbook for the late adopters/laggards — aligning business outcomes to use cases, aligning core practices for using AI to drive value across the organization, and retraining employees to prepare them for AI adoption. So, if you are driving a digital transformation journey, double down on AI — there is a ton of use cases and play books on applied AI today, lot more choice in capabilities and tooling, and a much better talent pool, than even a few years ago.

Enterprise AI/Machine Learning (ML) is a Hybrid Activity: Organizations that have cracked the AI nut, have realized Enterprise AI is a hybrid activity involving diverse teams from business analysts, to data scientists, data engineers, software developers, dev-ops and risk management functions. Realizing this is critical as I see organizations investing Auto-ML platforms (like Google Cloud Auto-ML, AWS Sagemaker, Azure AutoML, DataRobot, etc.) and expecting breakthrough results, only to highlight the gaps in investment in other areas. Auto-ML will automate a lot of mundane activity around ML, but you will still have to find other ways to access/prepare data, bring domain knowledge into your ML model development, assess risks. More on this in Part 2. Key is to invest in the entire Outcome to Activity process, investing in talent, and looking at organizational readiness across the entire value chain.

Solve for Data to Scale AI: Rob Thomas at IBM says more eloquently, “You can’t have AI without IA” (Information Architecture). Data is the oxygen in AI. With more and more of ML models and algorithms being readily available from the open source community and cloud marketplaces, it’s no surprise that the data used to train, optimize and power these models are the key to capturing value with AI. This makes Information Architecture (IA) more critical for AI. Having a robust data strategy, everything from understanding what data resides where, to accessing it and making available rapidly to data scientists and analysts will hold the key to successfully capturing value with AI. One feedback I heard from Chief Data Officers is they are usually knee-deep on Data Governance (more reactive, risk oriented) and are all orienting themselves towards more Growth oriented initiatives (Data-as-a-Service/Data Sharing, New Products/Offering insights, powering more AI initiatives)

Make Risk & Governance a key tenant of your AI Strategy: As usage has grown, so has awareness of the various risks of AI — from unintended bias to determining accountability. It’s critical to have a robust understanding of the strategic, operational, and ethical risks with AI. We have recently seen the uproar and increased negative sentiment on technologies like Facial Recognition (FRT) due to technical issues such as bias. It is one thing to look at a picture and say it’s a dog or a muffin, but a totally different legal and ethical issue to deny services to a racial minority due to the bias in your algorithmic decision making. Laws like the ones California and NYC are passing will become commonplace and regulatory frameworks like the one from EU will emerge . For organizations looking to deploy AI as a transformative force to serve customers, partners and suppliers, they will have to assess this from a risk and governance perspective — risk of reputation-loss, legal risk, compliance and regulatory risk etc. Although it’s still early days, auditability of AI systems should be an important part of the AI strategy, and it needs to go beyond model management and monitoring and into serving all of the stakeholders in the AI value chain. I’ll have more on this topic in the future.

Make AI a Corporate/ Enterprise Priority: In the 2020 O’Reilly State of AI in the Enterprise report a good chunk of the ~1400 respondents identified a lack of institutional support as the biggest problem in driving AI adoption, compared to the AI skills gap in 2019.

This is probably the most critical element that the market leaders have gotten right. This includes making sure there is end to end alignment across the organization on the need to invest in AI, to having the right business units support for driving AI projects, and building out a center of competence to make AI a lasting Entreprise capability. The other aspect of this is the realization that AI is still in its early innings in the universe, there is a lot of progress being made and there is a lot more to come. And as they say, if you want to go far, go as a team. The power of building and developing an ecosystem for your AI journey is critical. This includes partnering with research organizations, early stage startups working on cutting edge problems, the usual strategic vendors and partners. A great example is here how Anthem is transforming their business with AI, through innovation, ecosystem and enterprise leadership.

This is probably one of the most critical inflection points in our lives, both personally and professionally, and just like history shows us with all black swan (or gray-rhino if you prefer that) events this brings out about fundamental shifts in societies and businesses. AI has accelerated pretty rapidly due to commercialized research, availability of talent and cheaper compute and data. And right now presents a critical opportunity to transform or accelerate your organizational potential with AI.

Let me know what you think, share your stories and continue the conversation at @_ganeshp