AI News Roundup – Artist Rights Alliance letter, DALL-E model allows interactive editing, and more

Article co-written by Yuri Levin-Schwartz, Ph.D., a law clerk at MBHB.

To help you stay on top of the latest news, our AI practice group has compiled a roundup of the developments we are following.

    • The Opera web browser has announced that it now supports a number of downloadable large language models (LLMs) that execute locally, entirely within the browser. Some are general purpose and others are task specific. Their use is still experimental, so caution is warranted.
    • The Artist Rights Alliance has published an open letter asking that technology platforms “pledge that they will not develop or deploy AI music-generation technology, content, or tools that undermine or replace the human artistry of songwriters and artists or deny (artists) fair compensation for our work.” The letter is signed by over 200 musicians including Billie Eilish, the estate of Bob Marley, Chuck D, Diana Krall, Elvis Costello, Jon Batiste, Jon Bon Jovi, Marc Ribot, Nicki Minaj, Pearl Jam, REM, and others. The artists encourage the responsible use of AI while warning against its irresponsible use. On a similar note, instrument-maker Roland and The Universal Music Group have released a document titled The Principles for Music Creation with AI. The principles, while thin on detail, advocate for the protection of human-created works of music through copyright, that AI can be used to enhance human creativity, and that AI platforms be transparent.
    • OpenAI’s DALL-E model now allows interactive editing of generated images based on a user prompt. This feature can be accessed by way of the latest version of GPT4, which integrates ChatGPT and Dall-E. OpenAI has also announced that they are testing a voice generation tool that can mimic voices of individual humans. The tool’s initial uses are in the education and medical fields, but the company recognizes that this technology can be used to deceive and spread misinformation so its use comes with the appropriate caveats.
    • The creators of a fake posthumous George Carlin comedy routine have settled with Carlin’s estate, agreeing to take the content down, never upload again it to any platform, and not to use Carlin’s voice or likeness again without permission. Apparently, AI was used to generate at least part of the content but it is unclear what role human comedians played in the final product.