# Agentifying Agentic AI

Source: https://www.youtube.com/watch?v=2gajp32I2lQ
Recap page: https://rapidrecap.app/video/2gajp32I2lQ
Generated: 2025-11-29T19:04:15.818+00:00

---
## Quick Overview

The core argument presented is that current large language models (LLMs), despite their massive scale and flexibility, are fundamentally flawed because they lack explicit structure, social reasoning, and verifiable guarantees, making them unreliable for complex, real-world tasks compared to agentic systems designed with explicit coordination and accountability mechanisms.

**Key Points:**
- Current LLMs are fundamentally flawed because they rely on statistical inference from massive, unstructured data, rather than explicit reasoning or verifiable guarantees.
- The paper critiques the lack of structural coherence and social reasoning in LLMs, forcing them to rely on implicit guesswork, which leads to errors in complex tasks like booking flights or coordinating agents.
- The AMS (Agent-based Multi-Agent System) approach, exemplified by the proposed BDI (Belief-Desire-Intention) architecture, provides necessary guarantees for reliability, accountability, and coordination.
- The cost of using current LLMs for complex tasks can skyrocket (potentially hundreds of dollars a month) due to the need for extensive manual oversight to correct failures like mission creep or safety violations.
- A key insight is that intelligence does not need to be entirely contained within the agent; shared social context and explicit rules (like those in AMS) provide essential grounding.
- The authors suggest that future progress requires moving away from purely data-driven, monolithic LLMs toward hybrid systems that combine LLM flexibility with explicit, engineered structure for reliability and accountability.

![Screenshot at 00:08: The speakers introduce the central paradox that current LLMs, despite their power, are fundamentally broken because they lack the necessary structure and social reasoning for reliable, long-term goal achievement.](https://ss.rapidrecap.app/screens/2gajp32I2lQ/00-00-08.png)

**Context:** This video discusses the limitations of current large language models (LLMs) in achieving reliable, agentic behavior, drawing heavily from a research paper by Virginia and Frank Dignum which critiques the statistical nature of LLMs (like those based on large datasets) for real-world applications requiring accountability and coordination.

## Detailed Analysis

The video critiques the prevailing trend of building next-generation Artificial General Intelligence (AGI) solely on massive, flexible LLMs, arguing that this approach presents major paradoxes. The core argument, stemming from a paper by Virginia and Frank Dignum, is that these LLMs are fundamentally broken because they lack explicit structure, social reasoning, and verifiable guarantees needed for complex tasks. While LLMs excel at pattern recognition based on their training data (e.g., predicting the next word), they struggle with tasks requiring adherence to external rules or explicit intent, such as coordinating travel plans or managing multi-agent workflows. The presenters use the example of an agent booking flights where the LLM might choose a cheaper, indirect route via Rome, violating the explicit instruction to minimize travel time, because it lacks the social context and rule-following capability of a structured system. The paper contrasts this with the AMS (Agent-based Multi-Agent System) framework, specifically mentioning the BDI (Belief-Desire-Intention) architecture, which forces agents to define their internal thought processes and adhere to explicit social norms and rules. This structured approach provides crucial guarantees for reliability, accountability, and predictability, preventing issues like mission creep or costly failures when an agent attempts an action outside its defined scope. The conclusion is that the path forward for robust AI involves hybrid systems that integrate the flexibility of LLMs with the rigor of explicit, engineered structures to ensure coordinated, safe, and accountable behavior in the real world.

### Paradox of Current AI

- Massive race to build AGI using LLMs, but current systems are fundamentally broken due to lack of structure and social reasoning
- LLMs rely on statistical inference, not explicit reasoning or guaranteed behavior.

### Critique of LLMs

- Current LLMs are essentially black boxes based on statistical inference; they excel at pattern replication but fail to reliably execute complex, multi-step tasks requiring adherence to external rules or social context.

### The AMS/BDI Solution

- Agent-based Multi-Agent Systems (AMS) using frameworks like Belief-Desire-Intention (BDI) provide necessary structural reasoning, explicit intent tracking, and coordination rules.

### Practical Failure Example (Flight Booking)

- An LLM agent might book a cheaper, indirect flight (e.g., via Rome) when the explicit goal was shortest travel time, demonstrating a failure to adhere to implicit social/contextual norms.

### Cost of Unstructured AI

- Relying on unstructured LLMs for complex tasks leads to high costs (potentially hundreds of dollars a month) for human oversight to correct errors and prevent mission creep or safety violations.

### Future Path

- The viable path forward involves hybrid systems that combine the flexibility of LLMs with engineering rigor to enforce coordination, accountability, and adherence to shared social values and rules.

![Screenshot at 00:00: Opening screen featuring the 'Become A Member Today!' call to action over an oscilloscope-like display.](https://ss.rapidrecap.app/screens/2gajp32I2lQ/00-00-00.png)
![Screenshot at 00:34: Speaker discussing the core argument that current autonomous systems rooted in LLMs have fundamental flaws regarding coordination and long-term goals.](https://ss.rapidrecap.app/screens/2gajp32I2lQ/00-00-34.png)
![Screenshot at 01:36: Visual representation of the proposed solution: combining the power of an LLM with the ability to reason using tools like a calendar or code interpreter.](https://ss.rapidrecap.app/screens/2gajp32I2lQ/00-01-36.png)
![Screenshot at 02:28: Visual highlighting the core problem: forcing rigid structures onto flexible models leads to limiting intelligence, questioning the need for external guardrails.](https://ss.rapidrecap.app/screens/2gajp32I2lQ/00-02-28.png)
![Screenshot at 04:09: Speaker explaining the difficulty in verifying the output of purely statistical models, contrasting it with the need for explicit intent tracking.](https://ss.rapidrecap.app/screens/2gajp32I2lQ/00-04-09.png)