Staying updated in the fast-paced world of AI is no easy task. To help you keep track of recent developments in machine learning, as well as noteworthy research and experiments we may have missed, here’s a concise roundup.
Midjourney, an AI startup specializing in image and soon video generation, recently made changes to its terms of service related to IP disputes. This change reflects a shift towards more formal language, hinting at the company’s confidence in prevailing in legal battles over creators’ training data.
Generative AI models like Midjourney’s rely on a vast number of examples, sourced from public websites and repositories. While vendors argue for fair use in the training process, some creators dispute this, pointing to instances where models replicate training data.
Midjourney’s approach to handling copyrighted works has raised eyebrows, as the company has used works from various artists and popular franchises in its training data without proactive licensing agreements. This bold strategy could have significant legal implications if fair use is not upheld in court.
Aside from Midjourney, several other intriguing AI developments have surfaced recently:
AI-assisted ad controversy on Instagram: A director faced backlash for using another creator’s work without credit in a commercial.
EU scrutiny of AI platforms before elections: Tech giants are being questioned about their efforts to prevent election interference.
Google Deepmind’s AI for gaming: An agent trained on 3D game play can now perform tasks using natural language commands.
Exploring New Horizons in AI
AI models have revolutionized our understanding of molecular dynamics and properties, with tools like Microsoft’s ViSNet tackling complex molecular relationships.
AI’s impact extends beyond molecular science to diverse fields like COVID-19 variant prediction and atmospheric analysis. Researchers are leveraging AI to uncover hidden patterns and insights that were previously inaccessible.