OpenAI : Sora

About:

• Sora:

  • Sora is an AI model developed by OpenAI –– built on past research in DALL·E and GPT models –– and can generate videos based on text instructions.
  • Sora can create full videos in one go or add more to already created videos to make them longer.
  • It can produce videos up to one minute in duration, ensuring high visual quality and accuracy.
  • Features:
    • Sora can generate complex scenes with various characters, precise actions, and detailed backgrounds.
    • Not only does the model understand the user’s instructions, but it also interprets how these elements would appear in real-life situations.
    • It is capable of generating compelling characters that express vibrant emotions.
    • Sora can also create multiple shots within a single generated video that accurately persist characters and visual style.
  • Shortcomings of the model:
    • It may struggle with accurately simulating the physics of a complex scene, and may not understand specific instances of cause and effect.
    • For example, a person might take a bite out of a cookie, but afterward, the cookie may not have a bite mark.

• OpenAI

  • Establishment: December 2015 as a non-profit AI research organisation whose goal was to develop “artificial general intelligence,” or AGI.
    • AGI is essentially software that is as smart as humans.
  • The organisation said it wanted to ensure that AGI benefits all of humanity and no big tech company, like Google, would master the technology and monopolise its benefits.
  • Founding members: Among its founding members were Altman, Brockman, Reid Hoffman, the co-founder of LinkedIn, Amazon Web Services, Infosys, right-wing tech billionaire Peter Thiel and Elon Musk.
  • The evolution: Around two years after its inception, OpenAI released a report titled ‘Improving Language Understanding by Generative Pre-Training.’
  • This introduced the idea of Generative Pre-trained Transformers (GPTs).
    • GPTs are a type of large language model (LLM) that use transformer neural networks to generate human-like text.
    • GPTs are trained on large amounts of unlabelled text data from the internet, enabling them to understand and generate coherent and contextually relevant text.
    • They can be fine-tuned for specific tasks like: Language generation, Sentiment analysis, Language modelling, Machine translation, Text classification.
    • GPTs use self-attention mechanisms to focus on different parts of the input text during each processing step.
    • This allows GPT models to capture more context and improve performance on natural language processing (NLP) tasks.
      • NLP is the ability of a computer program to understand human language as it is spoken and written — referred to as natural language.

• ChatGPT

  • ChatGPT is a state-of-the-art natural language processing (NLP) model developed by OpenAI.
  • It is a variant of the popular GPT-3 (Generative Pertained Transformer 3) model, which has been trained on a massive amount of text data to generate human-like responses to a given input.
  • The answers provided by this chatbot are intended to be technical and free of jargon.
  • It can provide responses that sound like human speech, enabling natural dialogue between the user and the virtual assistant.

• Generative AI, or generative artificial intelligence:

  • Form of artificial intelligence (AI) in which algorithms automatically produce content in the form of text, images, audio, and video.
  • Generative AI is powered by foundation models (large AI models) that can multi-task and perform out-of-the-box tasks, including summarization, Q&A, classification, and more. 
  • These systems have been trained on massive amounts of data.
  • It works by using a Machine Learning (ML) model to learn the patterns and relationships in a dataset of human-created content. It then uses the learned patterns to generate new content. 
  • Typically, it starts with a simple text input, called a prompt, in which the user describes the output they want. Then, various algorithms generate new content according to what the prompt is asking for.
  • Bard, ChatGPT are popular Generative AI tools.