OpenAI admits having infringed copyright with ChatGPT: already preparing a retribution

GPT-5, Bard, and other picture-to-text alternatives like Midjourney rely on training from all sorts of sources, many of them copyright-related. Altman, CEO of OpenAIacknowledged in a meeting at the White House his concern about the use of copyright material.

[Cómo usar Google Bard en España, aunque la alternativa a ChatGPT no haya sido lanzada aquí]

The free will of ChatGPT, Bard and others

Sam Altman’s own words argue that they are working on new models in which if the AI ​​uses copyrighted content, it is paid or remunerated.

Altman was sparing in his declaration, and did not abandon the idea that he has for be able to pay these authors

And it is that several artists have taken legal action against Stable Difussion and Midjourney for the use of their protected content for learning their models for the generation of images from text.

It is more and more voices who put the cry to the sky in front of this misuse of content without anyone seeming to put a fence on it. Sarah Andersen and Karla Ortiz, among other artists, are the ones pushing for their voices to be heard and finally putting a limit to this inappropriate use.

Google polishes ethics with Firefly

The revenues generated by these companies are very high, but they do not come back to these artists and the ethics are left aside, even if Google begins to take its first steps.

Directly on Google I/O, Bard could be seen using Adobe’s Firefly to generate images from text. Adobe’s commitment is to the ethical use of materials copyrighted through its massive stock of images with Adobe Stock or with expired copyright.

Con esto, Google quiere asegurarse de que Bard pueda seguir encaminado a enfrentase a ChatGPT sin esos baches que pueden ser el uso indebido de contenido protegido. Un punto en el que OpenAI va a tener que mover ficha para que no le lluevan las demandas y ponga en peligro el dominio que ahora mismo está siendo su modelo de lenguaje ante otros.

También queda en el que se legisle debidamente para poner cerco a este libre albedrío que están siendo ahora mismo los modelos de lenguaje de diversas empresas. Es un primer paso que hay que dar, como mismamente China que forzó el uso de marcas de agua para que una imagen se supiera que ha sido generada por IA. 

Te puede interesar

Sigue los temas que te interesan