AI denied inventor rights in UK Supreme Court ruling

Pivot 5: 5 stories. 5 minutes a day. 5 days a week.

11. AI denied inventor rights in UK Supreme Court ruling

The UK Supreme Court has ruled that AI cannot be legally recognized as an inventor in patent applications, requiring the inventor to be a person. This decision concludes Dr. Stephen Thaler's dispute with the Intellectual Property Office over listing his AI, DABUS, as the inventor of two creations.

Despite Thaler's contention that DABUS autonomously generated the inventions, the court maintained that AI, as a non-person entity, does not have the legal standing to be credited as an inventor under the Patents Act 1977. The ruling has implications for the future of AI in the field of innovation and patent rights, reflecting ongoing legal and ethical debates surrounding AI's role in creative processes.

Read the full story here 

2. Rite Aid Faces 5-Year Ban on AI Facial Recognition

The U.S. Federal Trade Commission has banned bankrupt pharmacy chain Rite Aid from using AI facial recognition technology for surveillance for five years due to consumer harm. Between 2012 and 2020, the company's system erroneously flagged individuals as previous shoplifters.

This decision, pending bankruptcy court approval, follows a Reuters investigation revealing Rite Aid's deployment of the technology in predominantly lower-income, non-white neighborhoods. The ban reflects growing regulatory scrutiny over privacy and bias in AI surveillance practices.

Read the full story here 

3. Apple unveils AI breakthroughs for next-gen iPhones

Apple's latest AI advancements promise to transform iPhone experiences through two groundbreaking research papers. The first introduces HUGS, a technique to create animated 3D avatars from simple videos, offering up to 100 times faster training and rendering, paving the way for immersive telepresence and virtual try-ons.

The second addresses deploying LLMs on devices with limited memory, significantly improving inference latency, which could bring sophisticated AI assistants to consumer devices.

Read the full story here

4. Google's VideoPoet revolutionizes AI video generation

Google Research introduces VideoPoet, a LLM pushing the boundaries of AI-driven video generation. Diverging from traditional diffusion-based methods, VideoPoet employs transformer architecture to create longer, higher-quality video clips with consistent motion.

Trained on an extensive dataset, it can generate a wide range of outputs, including simulating camera motions and generating audio. Initial human raters preferred VideoPoet's output, highlighting its superior performance. While not yet publicly available, VideoPoet's potential for "any-to-any" generation tasks sets a new standard for video and audio creation in the AI space.

Read the full story here 

5. Child abuse material found in AI dataset, LAION-5B pulled

Researchers from the Stanford Internet Observatory discovered at least 1,008 instances of child sexual abuse material (CSAM) in the LAION-5B dataset, a significant resource for training AI image generation tools. This alarming finding could potentially enable AI models trained on this data to generate new instances of CSAM. LAION, the non-profit behind the dataset, has temporarily taken it down as a precautionary measure.

Despite efforts to filter illegal content, the vast scale of the dataset had already raised concerns. The discovery prompts serious ethical and legal questions about dataset curation and the responsibility of AI developers to ensure the safety and legality of their training materials.

Read the full story here