How AI Really Works - Intro to Open Source Large Language Models

Поделиться
HTML-код
  • Опубликовано: 11 сен 2024
  • Recording of a talk I gave on July 27th, 2024 at the Vancouver Public Library in Vancouver, Canada. It's designed to be an accessible introduction to AI, with the goal of demystifying Large Language Models (LLMs), using everything open source, so that people can start thinking of LLMs as just another technology, a tool in their toolkit, ready to be used in creative ways to various ends. It was delivered to an audience with diverse backgrounds, occupations & interests.
    Index:
    00:48 - About Me
    00:56 - Artificial Intelligence
    04:34 - Natural Language Processing
    07:15 - Large Language Models
    17:41 - Base Models
    19:54 - Next Word Prediction
    33:05 - In-Context Learning
    37:45 - Fine-Tuning Models
    38:54 - Instruction Tuned Models
    43:28 - Conversational AI
    51:29 - Summarization
    54:15 - Structured Data Extraction
    58:58 - Tool Use
    1:03:03 - Retrieval Augmented Generation
    1:05:35 - Q&A
    Audience Questions:
    23:16 - Are LLMs rule-based or purely predictive?
    27:27 - Are the big tech companies AI to train the AI models?
    39:48 - If the models are only predicting the next word, how is it doing all this intelligent work?
    47:58 - Are you using an actual live language model in this presentation?
    49:48 - Can I trust AI to answer important questions?
    56:48 - What are those weird triple backticks you're using, do you need that?
    1:06:25 - Is it silly to be polite to these models when prompting?
    1:08:28 - Which programming language(s) are you using?
    1:08:54 - Can you incorporate RegEx (Regular Expressions) into the process?
    1:09:31 - How close has this come to passing the turing test?
    1:10:38 - What are the privacy implications of using LLMs, when entering personal information into prompts?
    1:12:35 - If organizations wanted to ask questions of their large internal knowledge base, how would that work?
    1:15:47 - For personal use, what models do you prefer?
    1:18:32 - Are there any limitations to prevent people from using AI to do bad things?
    1:21:37 - Can you explain what Open Source is?

Комментарии • 6

  • @rudrabhattacharya3622
    @rudrabhattacharya3622 25 дней назад

    Wonderfull talk

  • @vaibhavtomer3440
    @vaibhavtomer3440 27 дней назад

    woww wonderfull video😍😍

  • @aydenrl
    @aydenrl 29 дней назад +1

    Kush is the goat!

  • @vaibhavtomer3440
    @vaibhavtomer3440 27 дней назад +1

    Hello bhaiya i am a recent Bachelor of Computer Applications (BCA) graduate with a strong passion for Artificial Intelligence and Machine Learning. I am eager to begin my career in this exciting field and am currently seeking an internship opportunity where I can contribute, learn, and grow. I would be grateful for any guidance or opportunities you might be able to offer as I take the first steps in my professional journey 🥺🥺🥺🥺🥺🥺

  • @vaibhavtomer3440
    @vaibhavtomer3440 27 дней назад

    🥰🔥🔥🔥