The 4 Horsemen of the AI-pocalypse – Why Enterprises Fail to Adopt AI

Partner Content

This author account is for Partner Content with Emerj advertising and content partnership clients. Learn more at

The 4 Horsemen of the AI-pocalypse

This is a contributed article by Ian Wilson, Founder at Strategy 4 AI – learn more about Ian’s online AI strategy courses here. Ian is also the former Head of AI for HSBC, one of the largest financial institutions in the world. To inquire about contributed articles from outside experts, contact [email protected]

As the use of AI to support business operations moves through its maturity cycle we have passed a number of key milestones along the way. However, following a pattern reminiscent of previous emerging technology introductions, organizations initially hired experts only to balk when they used blasphemous words like “infrastructure,” “industrialize” or “strategy” rather than soothing words like “use case,” “quick win” or “easy ROI.” 

Unfortunately, instead of sacrificing their misaligned expectations, many businesses sacrificed their experts… 

Fast forward a couple of years and many of those businesses, having attempted to cut corners and seeing mainly failure, now have a visceral understanding of what their experts were advising and are looking, with more experienced eyes, at how to move forward from this point.

However, many businesses are still making avoidable mistaken assumptions when it comes to the use of AI Capabilities to support business objectives. I like to call the most egregious of these assumptions The 4 Horsemen of the AI-Pocolypse:

1. Death by Use Case (The Use Case Horseman)

The assumption that an AI “Use Case” is a business process or a customer journey that they are not, in general. The problem with the word “Use Case” is that its definition changes depending on your perspective. Use cases can be viewed by business leaders as their internal business processes, by transformation leaders as customer journeys, and by solution vendors as solution capabilities. 

The AI use case, in its most general form, is a single task. This task is usually part of a business process that is, itself, part of a customer journey. However, and we are all guilty of this, we use the shorthand “use case” to apply to any and all of these cases. Business leaders, for example, could be pitched “AI for consumer loans” expecting that some software system can be dropped into their business that handles everything, end to end, when in fact the AI Capability is only deployed to perform a complex credit-scoring task, with the rest of the process and journey being someone else’s problem. This causes misalignment of expectations when we talk about “AI use cases”.

2. Death by Quick Wins (The Quick Win Horseman)

AI Project Expectation Matrix - Emerj Plus
Overpromising short-term ROI and measurable impact from “quick wins” is a common pitfall for project leaders. Image: Emerj’s AI Expectation Matrix, Source: Emerj Plus

The assumption that your teams have been sitting on their hands leaving the low-hanging fruit unpicked. This is rarely the case and AI is rarely quick or easy anyway.

Any large organization has transformation or project teams that have been working with a variety of automation technologies for the past decade with each of those technologies initially focused on “quick wins”. This leaves the landscape of AI suitable quick wins looking pretty barren.

3. Death by Budget (The Budget Horseman)

The assumption that budgets are only available for small, ad-hoc projects with no budget (or strategy) available to build the foundational components required to build scalable, industrial AI-based Transformation. In order to reduce the unit cost of AI Capabilities they have to be industrially scaled and deployed on platforms for the whole organization. This requires an initial investment that many organizations were unwilling to make while they focused on quick win, proof of concept, innovation use cases. Many, however, have seen the failure of that approach and are now receptive and proactive in building AI Capabilities in a strategic manner.

4. Death by ROI (The ROI Horseman)

The culmination of 1, 2 and 3 has been a widespread lack of ROI from AI-driven initiatives, which should not be surprising given what we have discussed. 

Many organizations started with the assumption that a few cheap, quick-win use cases could deliver easy benefits, equating AI ROI with RPA ROI with FTE reduction, and that was it. AI can, however, have far-reaching benefits beyond cost that need to be taken into account when calculating benefit / ROI, for example, increases in revenue, margin, customer experience and reductions in resource use, defects and risk. Add to those factors building internal skills and knowledge as well as reusable and potentially marketable data and model assets and benefit grows.

That being said, the current costs of deploying AI Capabilities are the primary killer of ROI and while organizations are focused on complex, customized, in-house deployments without moving to “AI as a Service” (AIaaS) that will not go away anytime soon.

Stay Ahead of the AI Curve

Discover the critical AI trends and applications that separate winners from losers in the future of business.

Sign up for the 'AI Advantage' newsletter:

Stay Ahead of the Machine Learning Curve

Join over 20,000 AI-focused business leaders and receive our latest AI research and trends delivered weekly.

Thanks for subscribing to the Emerj "AI Advantage" newsletter, check your email inbox for confirmation.