With my wife
Author: lechuck park
AI Infrastructure Architect & Technical Visualizer "Complex Systems, Simplified. I translate massive AI infrastructure into visual intelligence." I love to learn computer tech and help people by the digital.
infinite Gap
This illustration visually represents the philosophical exploration of the relationship between humans and technology, particularly AI. It emphasizes how technological advancements may narrow the gap between humans and machines, yet a fundamental difference will always persist. The concept of “reducing infinity” is depicted, showing that while AI can become more human-like, it can never be entirely the same. Ultimately, the image highlights that despite technological evolution, human judgment remains irreplaceable in final decision-making.
With ChatGPT
New Coding
The image titled “New Coding” illustrates the historical evolution of programming languages and the emerging paradigm of AI-assisted coding.
On the left side, it shows the progression of programming languages:
- “Bytecode” (represented by binary numbers: 0110, 1001, 1010)
- “Assembly” (shown with a gear and conveyor belt icon)
- “C/C++” (displayed with the C++ logo)
- “Python” (illustrated with the Python logo)
Below these languages is text reading “Workload for understanding computers” with a blue gradient arrow, indicating how these programming approaches have strengthened our understanding of computers through their evolution.
The bottom section labeled “Using AI with LLM” shows a human profile communicating with an AI chip/processor, suggesting that AI can now code through natural language based on this historical programming experience and data.
On the right side, a large purple arrow points toward the future concepts:
- “New Coding As you think”
- “With AI” (in purple text)
The overall message of the diagram is that programming has evolved from low-level languages to high-level ones, and now we’re entering a new era where AI enables coding directly through human thought, speech, and logical reasoning – representing a fundamental shift in how we create software.
With Claude
Mixture of Experts
This image depicts a conceptual diagram of the “MOE (Mixture of Expert)” system, effectively illustrating the similarities between human expert collaboration structures and AI model MoE architectures.
The key points of the diagram are:
- The upper section shows a traditional human expert collaboration model:
- A user presents a complex problem (“Please analyze the problem now”)
- An intermediary agent distributes this to appropriate experts (A, B, C Experts)
- Each expert analyzes the problem and provides solutions from their specialized domain
- The lower section demonstrates how this same structure is implemented in the AI world:
- When a user’s question or command is input
- The LLM Foundation Expert Model processes it
- The Routing Expert Model distributes tasks to appropriate specialized models (A, B, C Expert Models)
This diagram emphasizes that human expert systems and AI MoE architectures are fundamentally similar. The approach of utilizing multiple experts’ knowledge to solve complex problems has been used in human settings for a long time, and the AI MoE structure applies this human-centered collaborative model to AI systems. The core message of this diagram is that AI models are essentially performing the roles that human experts would traditionally fulfill.
This perspective suggests that mimicking human problem-solving approaches can be effective in AI system design.
With Claude
TCP/IP Better
This image is an informational diagram titled “TCP/IP and better” that explains various aspects of network protocols and optimizations.
The diagram is organized into three main sections:
- Connection
- Shows “3 way Handshaking” with a visual representation of the SYN, SYN+ACK, ACK sequence
- “Optimizing Handshake Latency” section mentions:
- QUIC (Developed by Google, used in HTTP/3) → Supports 0-RTT handshake
- TCP Fast Open (TFO) → Allows sending data with the first request using previous connection information
- Congestion Control
- Lists “tahoe & reno” congestion control algorithms
- Shows diagrams of Send Buffer Size with concepts like “Timeout 3-Dup-Ack” and “3-Dup Ack (Reno)”
- “Minimizing Network Congestion & Fast Recovery” section mentions:
- CUBIC → Less sensitive to RTT, enabling faster congestion recovery
- BBR (Bottleneck Bandwidth and RTT) → Dynamically adjusts transmission rate based on real-time network conditions
- Header Remove
- Shows TCP header structure diagram and “Optimize header” section
- “Reducing Overhead” section mentions:
- Compresses TCP headers in low-bandwidth networks (PPP, satellite links)
- Uses UDP instead of TCP, eliminating the need for a TCP header
The diagram appears to be an educational resource about TCP/IP protocols and various optimizations that have been developed to improve network performance, particularly focused on connection establishment, congestion control, and overhead reduction.
With Claude
Data Quality
The image shows a data quality infographic with key dimensions that affect AI systems.
At the top of the image, there’s a header titled “Data Quality”. Below that, there are five key data quality dimensions illustrated with icons:
- Accuracy – represented by a target with a checkmark. This is essential for AI models to produce correct results, as data with fewer errors and biases enables more accurate predictions.
- Consistency – shown with circular arrows forming a cycle. This maintains consistent data formats and meanings across different sources and over time, enabling stable learning and inference in AI models.
- Timeliness – depicted by a clock/pie chart with checkmarks. Providing up-to-date data in a timely manner allows AI to make decisions that accurately reflect current circumstances.
- Resolution – illustrated with “HD” text and people icons underneath. This refers to increasing detailed accuracy through higher data density obtained by more frequent sampling per unit of time. High-resolution data allows AI to detect subtle patterns and changes, enabling more sophisticated analysis and prediction.
- Quantity – represented by packages/boxes with a hand underneath. AI systems, particularly deep learning models, perform better when trained on large volumes of data. Sufficient data quantity allows for learning diverse patterns, preventing overfitting, and enabling recognition of rare cases or exceptions. It also improves the model’s generalization capability, ensuring reliable performance in real-world environments.
The bottom section features a light gray background with a conceptual illustration showing how these data quality dimensions contribute to AI. On the left side is a network of connected databases, devices, and information systems. An arrow points from this to a neural network representation on the right side, with the text “Data make AI” underneath.
The image appears to be explaining that these five quality dimensions are essential for creating effective AI systems, emphasizing that the quality of data directly impacts AI performance.
With Claude
LLM/RAG/Agentic
This image shows a diagram titled “LLM RAG Agentic” that illustrates the components and relationships in an AI system architecture.
The diagram is organized in a grid-like layout with three rows and three columns. Each row appears to represent different functional aspects of the system:
Top row:
- Left: “Text QnA” in a blue box
- Middle: A question mark icon with what looks like document/chat symbols
- Right: “LLM” (Large Language Model) in a blue box with a brain icon connected to various data sources/APIs in the middle
Middle row:
- Left: “Domain Specific” in a blue box
- Middle: A “Decision by AI” circle/node that serves as a central connection point
- Right: “RAG” (Retrieval-Augmented Generation) in a blue box with database/server icons
Bottom row:
- Left: “Agentic & Control Automation” in a blue box
- Middle: A task management or workflow icon with checkmarks and a clock
- Right: “Agentic AI” in a blue box with UI/interface icons
Arrows connect these components, showing how information and processes flow between them. The diagram appears to illustrate how a large language model integrates with retrieval-augmented generation capabilities and agentic (autonomous action-taking) functionality to form a complete AI system.
With Claude
Lechuck Park 