BUSINESS TRANSFORMATION

AI implementation – Hype vs. Reality

“Time Magazine, in its annual review of the year 2023, has named Sam Altman as the CEO of the Year and among the 100 most influential people across the globe.” Does this reference to OpenAI’s cofounder ring a bell? With just the mention of the name, Artificial Intelligence (AI) enthusiasts would know what we are talking about. Thanks to the exhaustive media coverage of what AI is and what it is capable of, there is a heightened awareness and eagerness across businesses to harness AI’s capabilities. There is, however, a sense of apprehension in deciphering where and how to start.

Through this article, we would like to simplify the nuances involved in bridging the gap between the aspirational and operational sides of AI leading to tangible business results.

Aspirational AI - Endless possibilities for business

Here are some aspirational changes AI could drive for business processes focusing specifically on transforming how work gets done:

1. Self-learning and adaptive workflows: Imagine AI that not only automates tasks but continuously learns and improves workflows. This self-learning AI would be a never-ending efficiency engine analysing processes in real-time to identify and eliminate inefficiencies. This would create a dynamic and constantly evolving business environment.
2. Zero-error and exception handling: AI could strive for flawless process execution eliminating human error and streamlining exception handling. Imagine AI proactively identifying potential issues within workflows, suggesting solutions, or even automatically correcting them.
3. Holistic process automation with human oversight: We aspire for AI that automates entire business processes, from initial triggers to final outputs, while maintaining human oversight for critical decision-making. This would free up human resources for higher-level tasks while ensuring control over sensitive areas.
4. Real-time process optimisation and collaboration: Imagine AI facilitating seamless collaboration between humans and AI agents across different processes. AI could analyse data in real-time, suggest improvements, and even dynamically allocate tasks between humans and machines based on expertise and workload.
5. Predictive process design and resource allocation: AI could move beyond simply automating existing processes. Imagine AI that can analyse your business goals and market trends, then intelligently suggest new, more efficient workflows, and allocate resources to optimise performance; akin to having a precognitive business advisor.

Operational AI - The reality check

Here are some observations on a few hindrances to achieving full AI potential:

1. Legacy systems- Many businesses cling to outdated legacy systems due to data security concerns, the high cost of migration, and the complexities of change management. These systems in turn end up being the Achilles' heel when it comes to the implementation of modern AI solutions.
2. As a domino effect, legacy business structures still rely on a lot of manual paperwork-based operations. Transforming these habits and workflows into a digital-first approach can be a long and arduous journey.
3. Narrow-focused solutions- Many current AI solutions focus on addressing isolated problems. This fragmented approach makes it difficult to feel a significant impact across the entire business value chain.
4. Uncertainty around model training guidelines- Uncertainty surrounds AI model training guidelines. Questions about how AI learns, the risks of data bias, and the ethical handling of sensitive information remain unanswered. Businesses often require customised model training processes, and a lack of global consensus on these issues creates roadblocks.
5. Finally, a significant hurdle to overcome is the lack of standardised tools to measure AI's short-term impact on business performance. This critical evaluation gap makes it difficult to demonstrate the true value of AI solutions, especially in areas where proven tools are scarce. This leads to uncertainty about ROI that hinders a wider adoption of potentially transformative technologies.

While some big questions are yet to be addressed, a clear path to AI adoption is emerging. The mantra for success lies in “start humble; keep it nimble.”

• Invest in pre-project design: Collaborative planning across business and transformation partners is the key to success. Designate an "AI Incubation Centre" within the organisation to assess feasibility and to build shared accountability. This upfront investment will define the project roadmap and set the stage for a smoother journey.
• Mitigate risks proactively: Before embarking on the AI journey, thoroughly evaluate risks related to data security, quality of outcomes, and potential biases. This transparency fosters trust and empowers teams for successful implementation.
• Focus on high-impact use cases: Avoid a "low-hanging fruit" with minimal user impact. Instead, prioritise addressing critical pain points that significantly affect a large user base. These strategically chosen use cases will deliver viable, visible, and measurable benefits.
• Start smart, scale gradually: Begin with a core use case and dedicate the necessary resources to achieve results. Avoid introducing too many changes at once. This staged approach ensures success and lays the groundwork for future scaling.
• Bridge the digital divide: A robust digital infrastructure is crucial for AI effectiveness. Establish proper IT integration with underlying applications to provide a seamless foundation for AI operations.

Unleashing the power of practical AI

While achieving a completely autonomous AI-run business is some distance away, the potential for AI to significantly improve operations is present and real. Consider a manufacturing company struggling with production delays and inefficiencies. Imagine integrating an AI solution that analyses real-time data from factory machines, pinpoints bottlenecks in the assembly line, and even suggests adjustments to optimise production flow. This is the true power of AI – not replacing humans but augmenting our capabilities and solving real-world business problems.

By following a practical approach that emphasises focused use cases, measurable results, and ongoing collaboration, businesses can bridge the gap between aspirational and operational AI. This measured approach coupled with continuous investment in AI talent and infrastructure will pave the way for a future where AI unlocks unprecedented levels of efficiency, productivity, and innovation.