DeepTech315: Apple & EU / Robotaxi Delay / AI Chatbot Research

Поділитися
Вставка
  • Опубліковано 2 лис 2024

КОМЕНТАРІ • 6

  • @riothecat
    @riothecat 3 місяці тому +5

    Gene and Doug, RAG is not the same as "training" an LLM. Retrieval-Augmented Generation (RAG) uses external information retrieval to answer questions. It first retrieves relevant documents, then the language model generates a response using both the retrieved data and its own knowledge. This enhances accuracy without needing to retrain the model.

  • @gr8dvd
    @gr8dvd 3 місяці тому +1

    Apple’s MO is to be ‘proactive’ in reaction to long simmering tensions with developers and/or regulators. Great discussion… especially liked the political questions put to AI models. Shows the relevance of the old adage GIGO - garbage in, garbage out… and value of good supplemental datasets.

  • @fib4invest14
    @fib4invest14 3 місяці тому +1

    It make sense Doug's answer re cahtbot. Grok training data is from X and it may be more lean to the opinion of majority of the people using X. It like raising a child in certain environment and his\her behavior is influenced by parents & people around him\her. 😊

  • @petez4608
    @petez4608 3 місяці тому

    Isn’t Gemini based on Claude?

  • @FrizzelFry
    @FrizzelFry 3 місяці тому

    It looks like Gene needs a vacation - take some time on the beach man