Cuestionario vinculado a una guía · Medio Nivel

Explicación de la demanda del New York Times v. OpenAI

Pone a prueba las afirmaciones principales del Times, la evidencia de regurgitación y la respuesta de OpenAI, y la pelea por el descubrimiento de los registros de salida de ChatGPT.

Rutas de guía relacionadasDemanda por derechos de autor entre Nyt y Openai
Pregunta 1 de 8

When and where was the New York Times lawsuit against OpenAI filed?

Vista previa de cada pregunta8 preguntas

Las respuestas y explicaciones aparecen a medida que realiza el cuestionario.

  1. When and where was the New York Times lawsuit against OpenAI filed?
    • January 2023 in California
    • December 2023 in federal court in Manhattan
    • June 2024 in Delaware
    • March 2025 in the UK High Court
  2. Why is Microsoft a defendant alongside OpenAI?
    • It wrote the Times articles
    • It owns the New York Times
    • It invested in OpenAI, provides its computing, and offers the models in products like Copilot
    • It was the judge's former employer
  3. What is 'regurgitation' in this case?
    • Models reproducing long passages of articles nearly word for word
    • OpenAI refusing to answer news questions
    • The Times republishing ChatGPT answers
    • Summaries that contain no original text
  4. How did OpenAI characterize the regurgitation examples?
    • As proof that it licensed the articles
    • As fabricated images
    • As intended product features
    • As the result of unusual, repeated prompting and a rare bug
  5. What did a preservation order in this case require OpenAI to keep?
    • Only its training code
    • ChatGPT output logs, including deleted conversations
    • The Times's subscriber lists
    • All Microsoft emails
  6. Why do retrieval and browsing features raise a different issue than training?
    • Because they never use any text
    • Because they are only available in Europe
    • Because they are covered by trademark law only
    • Because the copying happens at query time during use, not during training
  7. Which technique reduces memorized output without changing what the model learned?
    • Filtering outputs against known text
    • Retraining from scratch without the data
    • Increasing model size
    • Removing the tokenizer
  8. Which statement about the case's status is accurate according to the guide?
    • The Supreme Court ruled training is illegal
    • The Times withdrew all claims
    • It has not reached a final ruling on whether training is fair use
    • OpenAI was ordered to delete GPT-4