🤖
💭
👨🏻💻
Vibe Coding
AI Augmented Development
How did we get here?
🤖
💭
AI Augmented Development
How did we get here?
🤖
💭
Context is 👑
AI had been around
for a while...
🤖
💭
AI has existed in universities and
big tech organizations for many years.
A friend of mine studied neural networks at the Technion about 20 years ago…
Deep Blue beat Gary Kasparov in 1996
IBM Watson, Siri & Alexa are old news...
Interactive timeline from 2015 - 2025
A famous interview
🤖
💭
It was a futuristic, geeky gimmick.
It did not significantly impact our lives.
The machine is sentient!
🤖
💭
In 2022, Blake Lemoine, a Google AI engineer, was fired from Google for violating its employee confidentiality and data security policies after he publicly claimed its LaMDA artificial intelligence was sentient.
Later that year, we had the GPT moment…
GPT moment!
🤖
💭
In late Nov 2022, OpenAI made ChatGPT
available to the public for free!
100 million monthly users in 2 months!
It reached 1 million users
within just 5 days of its launch!
476 million December 2024
1 billion monthly users by October 2025
But Chat GPT was just a Front!
🤖
💭
While ChatGPT was offered for free
to the general public...
The web exploded with AI services as a result.
OpenAI also offered its AI models
via a paid API platform
Multi-Domain Use cases of AI models
🤖
💭
Generative AI
Text & Language Generation
- Conversational AI (Chatbots, Assistants)
- Content creation (emails, stories, summaries)
- Code generation
- Translation & grammar correction
Image Generation
- Creating & editing images with Midjourney, DALL.E, GPT...
Video Generation
- Video from text or images, Deepfakes & Avatars, Auto-edits, Animation
Audio Generation
- Text-to-speech (TTS), AI voice cloning, Music generation
Perception AI (Understanding the real world)
Vision (object recognition, tracking), Speech Recognition, Sensor Fusion (e.g. for robotics, drones)
Predictive / Analytical AI
Fraud detection, Forecasting, Diagnostics, Recommendations etc.
Tech Giants had to join in
🤖
💭
Google was forced to change its strategy and integrate AI into its search results to stay relevant.
Google also made its own models available via API, along with other public tools like Gemini and NotebookLM.
Microsoft, Amazon, Meta, Apple, and X have integrated their own AI models into their services and are offering them as cloud services as well.
New players like Anthropic and Mistral had emerged.
The Open Source Eco System
🤖
💭
Open source experienced a significant growth as well.
HuggingFace offers over 1M smaller, customized models that are free for download or used via their API.
Tools like Ollama, LM Studio, and OpenRouter make it easy to run models on your infrastructure (on prem).
Running models Locally require a strong infrastructure.
Why can't I use GPT or Claude for everything?
🤖
💭
Do you just want to use the model or include it in your product?
Is the LLM intended for general use?
Or will it need to be custom tailored for a specific use-case?
Can you use hosted LLMs over the network?
Or is "on-prem" a requirenment?
Some leading questions to help you choose a model that fit your needs
Does the model need to contain reasoning, or other traits?
Is the budget a consideration?
You may want to optimize for a faster customized model
Why do we need so many models?
LLM categories
🤖
💭
General-Purpose / Fast Model
Reasoning Models
Deep Search / Agentic Models
LLMs differ in architecture, training methods, and intended purposes.
They vary in speed, reasoning, and deep search capabilities.
1. General-Purpose / Fast Model
🤖
💭
Speed: High speed and low latency are primary goals,
achieved through smaller model sizes or specific optimizations.
Reasoning: Basic reasoning, may struggle with complex, multi-step logical problems without specific prompting (like Chain-of-Thought).
Deep Search: Generally do not have built-in deep search capabilities and rely solely on their internal training data.
These models are optimized for quick response times and cost-efficiency, making them suitable for everyday tasks like simple question-answering and content generation.
2. Reasoning Models
🤖
💭
Speed: Slower than general-purpose models,
often involves generating intermediate "thought" or "thinking" steps (Chain-of-Thought) to break down problems.
Reasoning: Highly capable in complex areas like mathematical proofs, scientific reasoning, logic puzzles, and debugging code, by systematically working through the problem.
Deep Search: Reasoning capabilities can be combined with search tools, but the core focus is on the logical processing of information, not information retrieval itself.
These models are engineered to excel at complex, multi-step reasoning tasks, often using sophisticated training techniques and architectural designs.
3. Deep Search / Agentic Models
🤖
💭
Speed: Slower, as they operate with a higher degree of autonomy, performing multiple steps over an extended period (minutes rather than seconds) to complete a comprehensive research task.
Reasoning: Utilize strong reasoning capabilities to comprehend user intentions, dynamically plan multi-turn retrieval, and synthesize information from various sources.
Deep Search: Involves extensive use of tools, like web browsing,
data analysis, and code execution, to perform deep information mining
and produce comprehensive reports.
These are advanced systems, often referred to as "search agents," that are designed for autonomous and in-depth information seeking and synthesis across multiple sources, typically the web.
How LLMs process text
🤖
💭
They operate by statistical probabilities.
While they may appear to understand, this is merely pattern matching, not human-like cognition.
LLMs do not "understand" text like humans do
Tokenization: converts text into tokens which can be
words, parts of words, or characters.
Embeddings: numerical representations that capture the meaning and context of these tokens.
Pattern recognition is the model's ability to find statistical relationships and structures within this numerical data.
Prediction: how the model determines the most statistically probable next word to generate coherent and relevant text.
Known issues when working with LLMs
🤖
💭
Training Data Limitations (a.k.a data cutoff)
Hallucination and Accuracy
Context Window Constraints
Reasoning Limitations
Inconsistency
Domain-Specific Limitation
Some of the solutions
🤖
💭
Web search tools, fact-checking tools
System prompts and Prompt engineering
Condense information before analysis
Break complex problems into smaller, sequential steps
Implement Retrieval Augmented Generation (RAG) systems
Implement human-in-the-loop workflows for critical decisions
Fine-tune models on domain-specific data when possible
Vibe Coding
By Yariv Gilad
