Sumit's AI & Tech Notes

Foundations of Generative AI

Generative AI

Working of Generative AI

Advantages of Transformer-based models:

Use case of Generative AI

Microsoft Generative AI solutions

  1. Microsoft Copilot
  2. Microsoft Foundry (Azure AI)
  3. GitHub Copilot
  4. Dynamics 365 Copilot
  5. Power Platform Copilot

Types of Generative AI Models

Challenges in Generative AI

  1. It can produce information that sounds correct but can be factually wrong or entirely fabricated. This challenge can be reversed by investing in fact-checking systems.
  2. It may produce different answers to similar prompts, making predictable performance difficult. Inconsistent outputs can disrupt workflows and decision-making.
  3. Bias occurs when AI reflect prejudices or imbalances presents in its training data. It introduces ethical and legal risks that can damage brand reputation and erode customer trust.
  4. It raises data privacy, energy consumption issues.

Understanding Prompt Engineering

Ground AI

Data readiness

Responsible AI Principles

  1. Fairness
  2. Reliability
  3. Safety
  4. Accountability

Machine Learning

Stages of Machine Learning lifecycle

  1. Define the task
  2. Collect and prepare the data
  3. Train and validate the model
  4. Deploy the model
  5. MLOps: Monitor and manage the lifecycle