OpenAI ya permite personalizar GPT-4o con datos propios para obtener un "mayor rendimiento"

OpenAI has introduced fine-tuning for GPT-4o, enabling developers to customize the model’s structure and tone using minimal training data. This capability allows for significant performance improvements across diverse domains, from coding to creative writing, by adapting the AI to specific user needs with greater efficiency and speed comparable to human response times. Apple is integrating ChatGPT into Siri and its writing tools, allowing users to leverage advanced AI capabilities directly on their devices. This collaboration enhances voice assistant functionalities and image generation while prioritizing user privacy by ensuring that requests are not stored and IP addresses remain hidden, fostering trust in AI-driven interactions. This development is highly relevant to open data, as it demonstrates the practical application of fine-tuning large models with proprietary or specialized datasets. It highlights the growing trend of adapting general-purpose AI to specialized tasks, emphasizing the importance of data privacy and the seamless integration of AI services into everyday consumer technologies, which shapes how open and private data are utilized in modern applications.

Source: 20minutos.es
Published on 2024-08-22