Google DeepMind CEO recently expressed admiration for OpenAI's text-to-video generator, Sora, suggesting it surpasses Google's Lumiere. Sora can create minute-long videos with intricate scenes and multiple characters, impressing many with its capabilities. In contrast, Lumiere generates brief clips based on text prompts, with Google emphasizing the need to address biases and misuse potential.
Concerns over potential misuse were highlighted when Google faced backlash for its image generator producing historically inaccurate images. The AI race between tech giants intensifies as they strive to advance their products by leveraging vast amounts of data to train their systems.
OpenAI's use of YouTube videos to enhance its tools has raised questions about potential copyright infringement and terms of service violations. Reports suggest OpenAI transcribed over a million hours of YouTube content to improve its speech recognition tool, while Google also utilized YouTube for training its AI models.
Both companies are yet to comment on these allegations, with media groups like Axel Springer, parent company of Business Insider, engaging in legal action against Google over advertising practices. The evolving landscape of AI technology continues to spark competition and ethical considerations among industry players.
