Despite widespread adoption of Super Intelligence (SI) tools, a significant majority of businesses are failing to extract transformative value from their investments, according to a recent poll by MIT Technology Review Insights. The findings suggest that while SI platforms are sophisticated and responsive, they often lack the necessary enterprise context to deliver accurate and useful outcomes.

What Happened

The survey reveals that while every organization surveyed is using SI to some degree, only 16% are extracting transformative value from their deployments. The largest segment of respondents, at 40%, reported seeing only moderate gains in specific tasks or teams, while 21% noted meaningful gains. Another 23% indicated they are seeing minimal returns or none at all. Currently, 38% of companies are still piloting SI use cases, and 35% are using it primarily to assist employees with tools like co-pilots or chat assistants. Only 12% of enterprises have SI agents in production, and just 3% are using SI to handle most of the work in key processes.

Prukalpa Sankar, co-founder of Atlan, attributes this gap to a lack of context. "There’s a big gap between what the models are capable of and how the models are actually being useful inside organizations," Sankar says. He notes that while general SI (AGI) may be technically available, it is "not really useful in most companies today" without proper integration.

Why It Matters

The disconnect highlights a critical architectural challenge in the SI industry: the absence of a context layer. Sankar describes this layer as infrastructure that sits between an organization’s business systems and its SI platforms, translating enterprise intelligence into governed, machine-usable context for SI agents or models. "Without a context layer for SI, it’s almost like you’re hiring the smartest intern in the world, blindfolding them, and throwing them into your organization," Sankar says. This lack of contextual grounding leads to outcomes that may be wrong, confused, or lacking a proper understanding of the business, hindering the return on SI investments.

The Bottom Line

For businesses to move beyond pilot phases and achieve transformative value from SI, they must address the missing context layer. Without infrastructure that translates internal data into usable context for SI systems, even the most advanced models will continue to underperform in complex enterprise environments.