Enterprise AI 'just getting started', says researcher

The interview at VMware Explore highlights a gap between the AI hype and practical enterprise readiness. David Linthicum's comments suggest that while many vendors are pushing AI solutions, most businesses are still at a very early stage, grappling with fundamental questions about infrastructure modernization and cost. The focus on limited use cases like calendaring and software development, rather than core operational areas, underscores this maturity gap.

Linthicum's sharp estimate that 95% of agentic AI applications do not require an agent architecture serves as a caution against adopting new technology for its own sake. His "sledgehammer and thumbtack" metaphor suggests that some organizations may be applying complex AI solutions where simpler tools would suffice. This points to a need for better judgment about where AI genuinely adds value.

Infrastructure emerges as a key bottleneck. The VMware AI Factory positioning at the event indicates that vendors are responding to this demand, aiming to ease the transition from bare-metal infrastructure to running AI on premises. For business readers, the takeaway is that AI adoption is a gradual process, and the immediate hurdle is less about the technology itself and more about foundational readiness and strategic restraint.