Three years after the widespread adoption of generative AI, businesses are finally seeing significant returns on their investments. Despite early skepticism and reports of pilot failures, AI is proving its value by transforming core business practices and driving operational efficiencies across various industries. This shift is moving beyond theoretical gains to measurable improvements.
Key Takeaways
- AI’s true value lies in leveraging proprietary data for tailored solutions.
- Stability and "boring" AI applications are crucial for long-term business success.
- "Mini-van economics"—designing AI systems around user needs and cost-efficiency—is key to avoiding financial pitfalls.
Unlocking AI’s Value Proposition
While initial AI adoption faced challenges, with many pilots failing to scale or demonstrate clear return on investment (ROI), a new understanding is emerging. Reports from institutions like MIT have highlighted these early struggles, leading some to question the viability of AI investments. However, experts now emphasize a shift in strategy, moving away from simply measuring AI’s ROI to understanding how it fundamentally enhances business operations.
The traditional approach to technology deployment involves cautious integration to avoid disrupting critical workflows. Introducing new technology, even if superior or cheaper, carries risks. The cost of data loss during a transition often outweighs the benefits of a new system, making stability a paramount concern for enterprises that rely on mature, long-standing technology.
The Power of Proprietary Data
A fundamental principle for realizing AI’s return is recognizing that "Your data is your value." While much discussion centers on AI model engineering, a widely adopted enterprise AI use case involves uploading proprietary business data into AI models. This practice narrows the AI’s focus, leading to faster, more accurate responses and reducing the need for extensive prompting.
This approach necessitates careful consideration of data governance and confidentiality. Enterprises must establish robust security measures and develop strategic negotiation tactics with AI vendors. Since vendors rely on non-public data to advance their models, companies can potentially exchange selective data access for services or price reductions. This creates a symbiotic relationship where businesses benefit from AI advancements, and vendors gain valuable training data.
Embracing Stability: "Boring by Design"
The rapid pace of AI model development, with hundreds of new models introduced annually, can destabilize business operations. When older, stable AI workflows are rendered obsolete by new model releases, businesses face disruption. Unlike the gaming industry, where users readily upgrade systems for new titles, business operations demand stability.
Successful AI deployments often focus on "boring by design" tasks – mundane but essential business processes. Automating tasks like legal or expense audits, while retaining human oversight for final decisions, exemplifies this approach. These applications don’t require constant updates to the latest models to deliver value. Abstracting business workflows from direct model APIs further enhances long-term stability, allowing for upgrades at a pace that suits the business.
"Mini-Van Economics" for AI
To avoid financial pitfalls, AI systems should be designed around user needs and consumption capacity, not just vendor benchmarks. Many companies fall into the trap of adopting new technologies based on supplier-driven metrics, rather than assessing their own operational capabilities and pace.
Instead of pursuing high-performance, costly solutions like "Ferraris," businesses should adopt "mini-van economics." This means designing for frugality by reconfiguring workflows to minimize spending on third-party services. For instance, customer support AI workflows can incur significant operational costs if not managed carefully. Companies that align AI system speeds with human reading pace (under 50 tokens per second) have successfully deployed scaled applications with minimal overhead.
Ultimately, the guidance for leveraging AI is to start practically, design for independence in technology components to ensure long-term stability, and capitalize on the fact that AI makes business data a valuable asset for both the enterprise and its technology suppliers.
Sources
- Finding return on AI investments across industries, MIT Technology Review.
Founder Dinis Guarda
IntelligentHQ Your New Business Network.
IntelligentHQ is a Business network and an expert source for finance, capital markets and intelligence for thousands of global business professionals, startups, and companies.
We exist at the point of intersection between technology, social media, finance and innovation.
IntelligentHQ leverages innovation and scale of social digital technology, analytics, news, and distribution to create an unparalleled, full digital medium and social business networks spectrum.
IntelligentHQ is working hard, to become a trusted, and indispensable source of business news and analytics, within financial services and its associated supply chains and ecosystems
