Artificial Intelligence (AI)

table of contents

• Artificial Intelligence (AI)
  • References
• AI for Java Developers
  • Glossary of Terms
  • LLMs and their Evolution
    • Building Bloacks of LLM and NLP
    • Evolution
    • Popular LLM Models
    • Advantages
    • Challenges
  • Prompt Engineering for Genrative AI
  • LLM Safety Guide

---

References

• Harvard CS50’s Artificial Intelligence with Python – Full University Course
• All Machine Learning algorithms explained in 17 min
• How JP Morgan Built An AI Agent for Investment Research with LangGraph - LangChain Interrupt
• O’Reilly
  • Build AI Apps with Spring AI, OpenAI, Ollama & SpringBoot

AI for Java Developers

• Enhance Problem Solving Abilities

Glossary of Terms

• LLM - Large Language Models
• NLP - Natural Language Processing : Enabling machines to understand, interpret and generate Human (Natural) Language.
• OpenAI -
• GPT - Generative Pretrained Transformer
• ChatGPT -
• RAG -

LLMs and their Evolution

• Large Language Model is a type of AI designed to understand, generate, and summarize human-like text by analyzing massive datasets using deep learning, specifically transformer architectures.
• LLMs work by predicting the next word in a sequence based on prompt context, enabling tasks like chatbot interaction, translation, and code generation.
• Populat LLMs
  • GPT models by OpenAI
  • Gemini by Google
  • Llama by Meta
  • Claude by Anthropic
• ChatGPT uses gpt-models of OpenAI behind the scenes

Building Bloacks of LLM and NLP

• Artificial Intelligence : computer systems designed to carry out tasks that normally need human intelligence
• Machine Learning : Focuses on building algorithms that allow the system to learn by itself.
• Deep Learning : Branch of ML focusing on algorithms that lead to Artificial Neural Networks. These excel inprocessing vast amount of data and are highly effective in tasks like image and speech recognition, as well as NLP
• GPT (Generative Pretrained Transformer) : Based on Transformer Architecture.

Evolution

• n-grams
• Neural Language Models (RNNs and LSTM)
• Transformer Architecture
• LLMs (GPTs)

Popular LLM Models

• Properiatary/Closed Source Models :
  • Not available for public download or modification
  • Paid services accessed through API calls
  • Examples
    • OpenAI - GPT
    • Gemini by Google
    • Claude by Anthropic
• Open Source Models :
  • Free available for Public
  • Examples
    • LLaMA by Meta
    • Mistral by MistralAI

Advantages

• Multilingual Capabilities
• Creativity and Content Generation
• Automation of Complex Task
• Rapid Prototyping

Challenges

• Bias and Fairness
• Ethical Concerns - Deepfakes
• Lack of Explainability - Industries like Legal, Medical or Financial
• Depends in Data Quality - Outdated data, biased data
• Knowledge CutOff
• Hallucination

!NOTE

Developers can now use the power of NLP to build many different kinds of applications

Prompt Engineering for Genrative AI

LLM Safety Guide