Skip to content

Commit

Permalink
Update index.md
Browse files Browse the repository at this point in the history
  • Loading branch information
KostyaCholak authored Oct 26, 2023
1 parent 1207f2f commit e013d51
Showing 1 changed file with 6 additions and 6 deletions.
12 changes: 6 additions & 6 deletions index.md
Original file line number Diff line number Diff line change
Expand Up @@ -10,11 +10,11 @@ nav_order: "1"

The AI landscape has been notably transformed by the emergence of Large Language Models (LLMs). Their rise has undeniably brought us closer to the horizon of Artificial General Intelligence (AGI). Yet, as with any innovative frontier, there are challenges that LLMs may not overcome.

#### 1.1. Controlling Probablity:
#### 1.1. Controlling Probablity

One of the nuances of LLMs lies in their control mechanics. While minor calibrations can be made to influence their outcomes, achieving a comprehensive grasp over their intricate functionalities is challenging. For instance, integrating a component as fundamental as long-term memory requires more than just a superficial adjustment. Existing methodologies like the RAG do offer potential solutions, but they are inherently limited and don't genuinely emulate organic memory dynamics. Altering the model's architecture is an option, but it mandates a comprehensive retraining process. And before we even delve into the complexities of alignment, it's evident that mastering the technical aspects alone presents substantial hurdles.

#### 1.2. Trainig Cost:
#### 1.2. Trainig Cost

Starting the training process of an LLM from the ground up is a resource-intensive endeavor. This doesn't just impact the exploration of newer architectural innovations but also presents a barrier to entry for independent and smaller-scale researchers. Hence, while LLMs present a tantalizing direction toward AGI, they also raise questions about access, equity, and diversity in research.

Expand All @@ -24,15 +24,15 @@ Yet, it's pertinent to acknowledge that the current trajectory of AI research he

The notion of Controllable AI extends beyond just creating a compliant machine. It's about understanding and delineating the balance between intelligence and controllability in artificial entities.

#### 2.1. Dealing With Ambiguity:
#### 2.1. Dealing With Ambiguity

One of the historic challenges algorithmic models grappled with was ambiguity. The idea is to craft an algorithm capable of processing ambiguity with the same depth and nuance as human cognition. By weaving in elements of context and real-world knowledge, the AI could potentially offer a richer, more comprehensive understanding of its environment.

#### 2.2. Conceptual Processing and Learning:
#### 2.2. Conceptual Processing and Learning

The evolution of AI hinges on its ability to process raw inputs—text, sound, images—and distill them into actionable concepts. The idea isn't just about reactionary protocols but building a system that learns and refines its understanding based on feedback. Such a system, rooted in pattern recognition and feedback loops, could potentially offer a more holistic understanding of its interactions and surroundings.
Our framework is based on the new pattern-matching techniques that can handle ambiguity. We process raw data (like text, audio, images) into a conceptual representation. It can then trigger different actions that are associated with the concept an agent was able to find in the raw data. For example a question might trigger an action of thinking about an answer and then telling it. But the agent's learning journey doesn't end there. By employing pattern matching on its subsequent actions and the feedback it garners from the environment, the agent is able to understand its influence on its surroundings, continiously building it's world model. This is made possible as both actions and outcomes are represented as graphs, allowing for a deeper, graph-based pattern recognition.

While this conceptual framework is ambitious and teeming with complexities, it underscores the potential pathways we could explore as we inch closer to AGI.
While this framework is ambitious and teeming with complexities, it underscores the potential pathways we could explore as we inch closer to AGI.

---

Expand Down

0 comments on commit e013d51

Please sign in to comment.