The wave of AI mania that has been spawned by OpenAI’s launch of ChatGPT in 2022 often seems to pivot around one tantalizing question: when will artificial general intelligence (AGI) arrive? In Silicon Valley, the narrative has focused on whether large language models will evolve into human-like superintelligence. But Prem Natarajan, Chief Scientist at Capital One, recently argued in SiliconANGLE, that debate is missing the real opportunity: Enterprise General Intelligence (EGI).
“Silicon Valley’s debates over the viability of human-like artificial general intelligence overlook the likelihood that true superintelligence won’t emerge in our phones but in enterprise systems.” — Prem Natarajan
So will large enterprise systems with their forest of applications, mountains of data, and thousands of documented processes provide a richer environment for advances in intelligence than what large scale but highly fragmented consumer adoption can provide? Instead of focusing on the hundreds of billions being spent chasing elusive and possibly distant AGI, do enterprises today have the chance to harness EGI to drive more immediate competitive advantage? It may be possible that because it doesn’t have to thrive in chaotic real-world settings, but only in the highly structured complexity of enterprise systems that EGI may offer immense returns far in advance of the realization of AGI.
What Makes EGI Different
Unlike consumer-facing AI, enterprise systems operate within human-designed environments where cause-and-effect relationships are engineered, not inferred. That makes them fertile ground for advanced AI. Supply chains, healthcare systems, retail networks, and energy markets all run on structured data and predictable workflows which provide exactly the conditions agentic AI needs to excel before requiring the superintelligence it may take thrive out in the wild unstructured world of consumers.
EGI differs from AGI in three fundamental ways:
- Structured Environments: AGI would need to navigate the full unpredictability of the real world. EGI leverages enterprise systems with built-in rules, guardrails, and data structures.
- Agentic Autonomy: Today’s foundation models are reactive, waiting for human prompts. EGI emphasizes agentic AI that can set goals, plan actions, and execute them across systems without constant supervision.
- Practical Feasibility: While AGI remains largely speculative, EGI is tractable. The building blocks of autonomous agents, real-time orchestration, and enterprise integration are already being deployed today.
As Natarajan notes, the leap to EGI is less about bigger models and more about orchestrating intelligence across platforms:
“The critical leap to EGI isn’t better language models but agentic AI that can set goals, plan actions and execute them across enterprise systems without constant human oversight.”
The Five Capabilities of EGI
According to Natarajan, EGI requires five core capabilities (with one underlying enabler):
- Autonomous goal decomposition and planning
- Persistent awareness across heterogeneous systems
- Multi-system action selection and verification
- AI-ready enterprise ecosystems (data pipelines, APIs, integration)
- Compositional reasoning that unifies conflicting terminologies
- Continuous reinforcement of learning over time
These may sound futuristic, but pieces of this are already visible. Salesforce has proposed an Enterprise General Intelligence readiness framework that emphasizes benchmarks, guardrails, and reliability testing for AI agents. Others have begun experimenting with large action models—systems designed not just to generate text but to plan and execute workflows across multiple applications. At NewTide we are implementing these solutions today that act across multiple systems with clear objectives and measurable accuracy and reliability.
Why EGI Will Emerge Before AGI
I believe that as the term is being defined, the answer is yes. EGI can be accomplished with the current level of LLM capability. With the well-structured and data rich environments that corporate enterprises already run today we can rapidly advance agentic intelligence that is proactive and mission driven. This situation provides both technological and commercial drivers for EGI.
- Technological: Orchestrating multiple enterprise systems is a hard but bounded problem that is far more tractable than replicating human-level reasoning across every domain of life. Each company is complex but also finite.
- Commercial: The business case is immediate. Companies that succeed in deploying EGI won’t just automate tasks they are creating autonomous business intelligence that knows more than any single employee and increasingly can act in real time.
“Companies that achieve EGI-level capabilities won’t just build better software; they’ll create autonomous business intelligence that knows more than any single human and can act on that knowledge in real-time.” — Natarajan
For industries like energy and convenience retail, the potential is profound. Imagine an EGI-enabled enterprise where disruptions in fuel supply chains are autonomously rerouted, convenience store labor schedules are optimized against real-time inputs of worker availability and store level demand, or emissions compliance reporting is generated and validated without human intervention.
RisingTide: A Pathway to Enterprise General Intelligence
At NewTide, we see RisingTide as a practical foundation for companies ready to begin their journey toward EGI.
- Agentic Autonomy: Through RisingTide’s Shipyard workflow builder, enterprises can create AI agents that plan, act, and learn across multiple systems—mirroring the autonomy at the heart of EGI.
- System Integration: With Data Helm, RisingTide provides the connective tissue for enterprise ecosystems—structured data pipelines and API orchestration that are prerequisites for EGI.
- Governance and Reliability: Built-in compliance checks, logging, and error-handling mean RisingTide agents operate within guardrails, ensuring the “trust layer” that Natarajan and others emphasize.
- Human-AI Collaboration: Just as Natarajan envisions humans as arbiters and auditors of EGI, RisingTide is designed to augment, not replace, decision-makers in our highly specialized supply chains and operating locations.
The Energy and Convenience Sector: A Natural Proving Ground
The energy and convenience sector is well positioned to lead in EGI adoption:
- Structured Complexity: Fuel logistics, retail pricing, labor scheduling, loyalty programs, and environmental compliance all run on structured systems that provide an ideal proving ground for EGI.
- Operational Impact is Immediately Meaningful: Margins are tight, volatility is high, and efficiency gains have immediate bottom-line impact.
- Data Richness: From real-time fuel pricing that updates by the second to inventory levels changing every minute down to millions of customer transactions everyday, the sector is data-rich and increasingly API-enabled.
RisingTide’s deployments in energy and convenience already point in this direction: automating bill of lading workflows for fuel distribution, embedding AI-native CRM intelligence for sales, and orchestrating multi-agent processes across both trading and retail operations.
Conclusion: Quietly, the Future Is Arriving
I believe the signal among all the noise and hype around viral consumer apps or debates about artificial consciousness will likely come from enterprises deploying truly autonomous systems that act intelligently in business operations.
“That transformation is closer than most realize, and its impact will be far more immediate than any theoretical AGI breakthrough.” — Natarajan
For companies in the energy and convenience sector, the question is not whether to adopt AI, but whether to build toward true Enterprise General Intelligence. This takes time and those that start early with what I call “useful but boring” automation steps that provide the immediate payback that then drive self-funded follow on investments will have a lead that is not easily overcome by throwing money or resources at the opportunity later. With RisingTide, NewTide offers a pathway to that future: where enterprise intelligence doesn’t just assist, but acts, learns, and collaborates to redefine what’s possible.
References
- Natarajan, Prem. The AGI debate is misguided. The bigger potential is Enterprise General Intelligence. SiliconANGLE, 2025.
- Salesforce, CIO.com, and Diginomica coverage of the EGI framework and agentic testing.
- McKinsey & TechRadar analysis of the “Enterprise AI Paradox.”
