NEWS IN BRIEF: AI/ML FRESH UPDATES

Get your daily dose of global tech news and stay ahead in the industry! Read more about AI trends and breakthroughs from around the world

Unbiased AI: Accuracy Preserved

MIT researchers developed a new technique to improve machine-learning model accuracy for underrepresented groups by removing specific data points. This method addresses hidden biases in training datasets, ensuring fair predictions for all individuals.

The Persuasive Power of AI in Video

OpenAI's new tool, Sora, creates realistic video clips from prompts, raising concerns about the blurring line between reality and AI-generated content. Despite impressive visuals, witnessing the uncanny realism left the journalist feeling more melancholic than amazed.

Enhancing Presentations with Multimodal Foundation Models

Two approaches to gain insights on multimodal data: embed first, infer later with Amazon Titan Multimodal Embeddings, and infer first, embed later with Anthropic’s Claude 3 Sonnet. Evaluation using SlideVQA dataset, providing concise responses to user questions.

Streamline Enterprise Automation with Amazon Q Business Plugins

Amazon Q Business uses generative AI to boost employee productivity with real-time data access and seamless integration with enterprise apps through plugins. Employees can perform actions and access non-indexed data across multiple applications using natural language, streamlining workflows and saving time.

Sora: OpenAI's AI Video Generator Now Accessible in US

OpenAI's Sora, a text-to-video generator, now available to all in the US, creates AI video clips from written prompts. Users can see their prompts come to life, like a family of woolly mammoths in an open desert, through Sora's innovative technology.

Simplify Data Quality

Summary: Learn three zero-cost solutions to improve data quality efficiently. Utilize old-school database tricks, create custom dashboards, and generate data lineage with Python. Simplify processes and reduce complexity for better data quality outcomes.

Introducing Pixtral 12B on Amazon SageMaker JumpStart

Pixtral 12B, Mistral AI's cutting-edge vision language model, excels in text-only and multimodal tasks, outperforming other models. It features a novel architecture with a 400-million-parameter vision encoder and a 12-billion-parameter transformer decoder, offering high performance and speed for understanding images and documents.

Maximizing LLM Performance: The Power of Hybrid Models

Implementing Speculative and Contrastive Decoding enhances text generation quality and efficiency using large and small language models. Contrastive Decoding prioritizes tokens with the largest probability difference between models for high-quality outputs.

AI Saves Lives: Reducing Stillbirths in Malawi

The only hospital in the country using foetal safety software reduced baby fatalities by 82% in 3 years. A routine ultrasound at Area 25 health centre in Malawi's capital saved Ellen Kaphamtengo's baby from potential asphyxia.

Enhancing GPS Accuracy: Mapping Kinematic Solutions

Geospatial datasets use dead reckoning to infer GPS locations between known points, impacting data quality and value. Telematics data from vehicles includes various signals like GPS location, speed, and more, with implications for data handling and analysis.