← return to index
imagoresearch.in

Defining General Executable Intelligence: The Shift from Conversational AI to Executable Logic.

Author: Mehboob Hassan
Published: March 2026
Protocol: imago.chat core architecture
Abstract

For the past half-decade, human-computer interaction has been bottlenecked by the paradigm of "conversation." Large Language Models (LLMs) have been treated as highly articulate search engines or digital advisors. This paper introduces General Executable Intelligence (GEI)—a deterministic framework where human intent bypasses the conversational user interface entirely, compiling directly into financial, legal, and spatial protocols to render immediate reality within a unified ecosystem.

01. The Chat Bottleneck and Linear Friction

Since the advent of generalized Large Language Models, the technology industry has normalized a fundamentally inefficient architecture: humans typing commands into a chat window, waiting for linear text generation, and then manually carrying out the actions described by the generated text.

Early research into LLM agency, such as the ReAct framework, demonstrated that models could be prompted to synergize reasoning and acting.[1] However, these systems were still fundamentally constrained by external Application Programming Interfaces (APIs) and the latency of third-party platforms. If a creator wanted to host an event, the AI could write the promotional copy, but the creator was still forced into a manual "Copy/Paste" workflow—navigating to an external ticketing platform, creating an account, and configuring the pricing logic.

This separation between generation (the AI's text output) and execution (the human's labor within a GUI) represents the "Chat Bottleneck." It artificially restricts artificial intelligence to the role of an advisor. The intelligence is locked behind a read-only interface, failing to achieve true spatial utility.[2]

Fig 1.0: System Architecture Comparison System: ONLINE
Legacy LLM Pipeline
1. Prompt Ingestion Type: String | Latency: 200ms
2. Text Generation Output: Natural Language | 15 tokens/s
3. Human Bottleneck Action: Manual GUI Interaction / Copy-Paste
Imago GEI Framework
1. Intent Parser Type: NLP -> AST | Latency: 45ms
2. State Compiler Output: Executable JSON Payload > await Router.dispatch()
tix.imago.chat
[ Commerce ]
ledger.imago.chat
[ Finance ]
pact.imago.chat
[ Legal ]
vault.imago.chat
[ Routing ]

02. The Architecture of Executable Payloads

General Executable Intelligence (GEI) operates on a distinct premise: the optimal output of an intelligence system is not conversational text, but deterministic state mutation. Research by Schick et al. (2023) regarding Toolformer proved that language models could teach themselves to use external tools via API calls.[3] GEI takes this to its logical conclusion: the AI does not just call external tools; it is the infrastructure.

By establishing a unified spatial environment—architected natively within the imago.chat ecosystem—we eliminate the latency of third-party handoffs. The system parses natural language directly into executable JSON payloads that trigger autonomous endpoints instantly.

The Autonomous Imago Nodes

Within the Imago framework, we have engineered critical nodes of execution that bypass traditional web interfaces entirely:

1. Spatial Commerce (tix.imago.chat)
Instead of navigating a dashboard to sell access, a user simply commands, "Drop 50 tickets for my Guwahati show this Sunday at ₹499." The GEI instantly compiles a secure, active checkout environment at tix.imago.chat, ready to process UPI payments. The intent is translated into commerce in under two seconds.

2. Financial Logic (ledger.imago.chat)
Invoicing and billing represent massive cognitive friction. Through GEI, commanding the system to "Bill the agency ₹1,50,000 for creative direction" automatically calculates the necessary GST splits, drafts the invoice payload, and stages it on the ledger.imago.chat node for immediate dispatch.

3. Legal Deployment (pact.imago.chat)
The generation of binding agreements is handled autonomously. The intelligence understands contextual scope, retrieves parameters from previous interactions, and generates secure service agreements waiting for cryptographic signatures at pact.imago.chat.

03. The Metamorphosis to State-Engines

We must accept that applications, as we have known them for the last fifteen years, are merely visual, manual wrappers for database commands. The era of the fragmented "App" is effectively ending, giving way to the era of the "Unified Protocol."

Chat was merely the cocoon. Executable Intelligence is the metamorphosis. The next iteration of the internet is not a destination you browse; it is an environment that acts on your behalf. imago.chat represents the architectural consciousness that will drive this post-web creator economy—transforming absolute friction into absolute flow.

References

[1]
Yao, S., Zhao, J., Yu, D., Du, N., Shafran, I., Narasimhan, K., & Cao, Y. (2022). ReAct: Synergizing Reasoning and Acting in Language Models. arXiv preprint arXiv:2210.03629.
Read Paper ↗
[2]
Victor, B. (2014). The Humane Representation of Thought: A Trail Map for the 21st Century. Dynamicland / Communications of the ACM.
View Research ↗
[3]
Schick, T., Dwivedi-Yu, J., Dessì, R., Raileanu, R., Lomeli, M., Zettlemoyer, L., ... & Scialom, T. (2023). Toolformer: Language Models Can Teach Themselves to Use Tools. arXiv preprint arXiv:2302.04761.
Read Paper ↗