AI Disillusionment and the State of the Market
by Max Vermeir, Senior Director, AI Strategy
I recently had the opportunity to attend a fascinating session exploring generative AI, delivered by Neil Ward-Dutton, VP of Automation, Analytics & AI at IDC.
It felt a bit like stepping onto a Class M planet in the Star Trek Universe: an initial sense of anxiety as we prepared for the unknown, followed by a relieving feeling of familiarity as we encountered truths that we’ve encountered before. Neil didn’t just skim the surface of this generative AI wave; he dove deep, unravelling the hype and offering practical advice on how businesses can (and should) draw value from this tech. Here’s a play-by-play recap from my perspective.
The generative AI hype vs. reality
Generative AI has been the talk of the town for about two years now, sending ripples (and rogue waves) through industries. The hype is irresistible—everything from automated text generation to customer service chatbots has been hailed as transformational. However, much of this remains hypothetical, like a lightsaber without kyber crystals to power it.
Neil framed it well: many businesses are still wandering aimlessly through an AI-saturated market, overwhelmed by flashy promises but unsure about finding or integrating the right solutions. Yet, despite the ambiguities, here’s why we can’t dismiss it outright: according to IDC data, 45% of businesses believe generative AI will impact their operations within 18 months, equating it to tectonic forces like ERP and robotic process automation.
But the million-dollar question persisted throughout the presentation: How do we get from "wow, look what ChatGPT can do" to tangible, actionable results? The key is a Jedi-like focus on use cases, clarity, and strategy.
From hype to practicality: bridging the gap
Now, here’s where things got interesting—over two-thirds of organizations are investing in generative AI in some form, with only 9% abstaining completely. But the catch? A lot of this budget goes into proof-of-concept (POC) projects, averaging 34 per organization annually. Some businesses are running over 100 POCs, essentially adopting a “throw spaghetti at the wall and see what sticks” strategy.
FOMO—the fear of missing out—is a big driver here. Executives, eager to showcase an innovative AI narrative to stakeholders, often scramble for quick implementations without a clear sense of purpose. Picture your CEO wielding generative AI like Tony Stark testing a new Iron Man suit—flashy, but often misaligned with practical needs.
Instead, successful businesses follow a roadmap grounded in seven pillars: strategy, governance, people, application, platform, data, and infrastructure. Neil emphasized a vital point: only some pillars are tech-focused. AI success hinges just as much on business transformation as it does on technology. It’s about syncing executive-level visions for AI with boots-on-the-ground processes.
That said, not all use cases are created equal. The presenter categorized them into three buckets:
- Individual productivity boosts: Low barrier to entry, but easy for competitors to replicate.
- Transforming products/services: Aspirational goals requiring deep alignment with culture, policy, and strategy.
- Process and functional value: Here lies the golden snitch of AI’s impact, offering measurable ROI through integration across specific business functions like procurement or legal.
The third category, he explained, is where the clearest path to success lies—connecting technology with real, tangible business impact.
AI as automation’s new hero
If AI and automation tools form an Avengers-like team, generative AI plays the techie tie-wearing Bruce Banner rather than the Hulk-smashing bombastic disruptor many assume it to be.
Neil broke down automation workflows into a loop, and this stuck with me as one of the most eye-opening parts of the event. Here’s the loop in a nutshell:
- Discover processes. Unearth what’s really happening within workflows.
- Understand. Get granular insights into bottlenecks and inefficiencies.
- Design improvements. Reimagine a better way to work.
- Deploy. Roll out the optimizations and monitor outcomes.
- Continuously adapt. Refine and enhance based on new insights.
Each part presents what Neil called “ripe fruit for picking” when it comes to AI’s role in automation. Process AI and Document AI were highlighted as on-ramps to this strategy. Process AI tools analyze everything from procurement approvals to sales handoffs. Document AI, meanwhile, goes a step further—extracting actionable insights from unstructured data like contracts or memos.
This was a great reminder that AI best works hand-in-hand with humans driving structure, measurability, and trust.
Cutting through hype and charting the course
One of the slickest takeaways of the session came from addressing disillusionment, something that’s inevitable for businesses going the “license first, purpose later” route with AI adoption. This is like firing a phaser randomly and hoping it hits the right target—it’s bound to disappoint (or worse).
The ultimate key to avoiding this is a focus on transparency, trust, and measurable value. By anchoring investments in areas where organizations already have critical visibility (like processes ripe for optimization), businesses stand a better chance at building repeatable, scalable success stories in AI.
Final thoughts—a practical AI odyssey
Generative AI doesn’t promise quick fixes, but it offers immense potential for those who apply it with purpose.
If you’re venturing into the AI galaxy, it’s time to go beyond curiosity and get clear on your strategy. Start with defining measurable use cases. Leverage tools like Process AI and Document AI to translate POC experiments into something sustainable. And remember—this isn’t a one-and-done transformation. AI needs to live within the fabric of your business, evolving alongside it.
To put it plainly, don’t just wield generative AI like the latest gadget in a sci-fi blockbuster—make it the engineer rebuilding your starship’s warp core for maximum efficiency.