La primera alternativa de código abierto a ChatGPT ya está lista: así es StableLM
Stability AI has launched StableLM, an open-source conversational model designed to compete with closed alternatives such as ChatGPT. This development demonstrates that smaller models can achieve high performance through proper training and diverse datasets. Its public availability enables the technical community to generate both text and code, although initial quality may vary without further fine-tuning. The relevance to open information lies in Stability AI’s commitment to transparency. By releasing the models, the company aims to foster trust and allow researchers to audit the models’ internal workings. This facilitates the identification of risks and the development of safety safeguards that would be impossible to achieve with proprietary approaches, thereby strengthening democratic control over the technology. Although there are concerns about potential malicious uses, such as phishing, the company maintains that open source is the right path. The release of these academic and research-oriented tools seeks to overcome the limitations of closed models. Ultimately, this move reinforces the idea that technical openness is essential for ensuring safer, more interpretable, and more ethical artificial intelligence systems.
Source: 20minutos.esPublished on 2023-04-22