Summarized by: Liam Rodriguez [medicalxpress.com]
Researchers at North Carolina State University have developed an AI and computer simulation method to train robotic exoskeletons, enhancing human mobility by reducing energy expenditure during activities like walking, running, and climbing stairs. This machine-learning framework eliminates the need for extensive human training and handcrafted control laws. In tests, participants using the exoskeletons showed significant reductions in metabolic energy consumption: 24.3% less for walking, 13.1% less for running, and 15.4% less for climbing stairs. The method aims to rapidly advance and broadly apply assistive robots, including those for mobility-impaired individuals.
BlinqIO raises $5M to automate software testing using generative AI
Summarized by: Liam Rodriguez [siliconangle.com]
BlinqIO Inc., a startup specializing in generative AI for software testing, has secured $5 million in funding led by Flint Capital, with participation from Inovia Capital’s Precede Fund I. BlinqIO has developed an AI test automation engineer that autonomously creates and maintains test scripts, potentially replacing human test engineers. This tool, powered by a proprietary large language model, supports over 50 languages and aims to alleviate the software testing bottleneck, allowing developers to focus on higher-value tasks. The global software testing market is projected to exceed $15 billion by 2027, highlighting significant growth opportunities for BlinqIO.
Summarized by: Liam Rodriguez [developer.nvidia.com]
NVIDIA has set new records in generative AI performance and scale with the MLPerf Training v4.0 benchmarks. Their advancements span various applications, including code writing, storytelling, music composition, image creation, and video production. Additionally, NVIDIA is enhancing large language models (LLMs) with Retrieval-Augmented Generation (RAG) to accelerate enterprise AI. Their AI foundation model, VISTA-2D, is advancing cell segmentation and morphology analysis in genomics. NVIDIA’s H200 Tensor Core GPUs and TensorRT-LLM have also set MLPerf LLM inference records. Other notable developments include innovative training techniques for diffusion models and applying Mixture of Experts (MoE) in LLM architectures. NVIDIA continues to push the boundaries of AI, making significant strides in both performance and practical applications.
New breakthrough makes Generative AI economically viable for legal teams
Summarized by: Liam Rodriguez [www.jdsupra.com]
A new breakthrough by Hanzo makes Generative AI economically viable for legal teams. Hanzo’s innovative engineering reduces data analysis and review costs by a factor of ten. Instead of relying on large language models (LLMs) like Chat-GPT, Hanzo’s Spotlight AI uses smaller models hosted on secure private cloud instances, optimizing performance and cost. The GB-based pricing model, rather than the token-based one, significantly lowers costs, making AI adoption feasible for legal teams. Efficient management ensures the LLM operates only when needed, aligning costs with usage. This approach opens new use cases in data discovery and compliance while reducing data discovery time by 90%.
Summarized by: Liam Rodriguez [www.thedailystar.net]
Previous headlines:
Adobe has integrated generative AI into Photoshop, allowing users to create images from simple text prompts, similar to Midjourney and DALL-E. This makes Photoshop more accessible to untrained users, simplifying complex editing tasks. For instance, a demo showcased the creation of a detailed image using a four-word prompt, significantly reducing editing time. However, there are concerns about job displacement due to AI. To address this, Adobe introduced Content Credentials, a tool to verify AI-altered images, and adheres to ethical practices by training its AI on licensed images and compensating contributing artists.
Microsoft, NVIDIA Partner on AI Hardware, Extend Reach of Copilot+ PCs
Summarized by: Aisha Patel [accelerationeconomy.com]
Previous headlines:
Microsoft and NVIDIA have announced a partnership to enhance AI hardware integration, specifically focusing on Copilot+ PCs. NVIDIA’s latest RTX AI PC laptops from Asus and MSI will include the Copilot+ experience, with a free update for Windows 11 AI PCs to support these features. The collaboration also involves developing the Windows Copilot Runtime, enabling generative AI capabilities in both native and web applications on Windows. Although a specific release date hasn’t been disclosed, this partnership underscores the strong alliance between Microsoft and NVIDIA in advancing AI hardware and software integration.
Productionize Foundation Models from SageMaker Canvas
Summarized by: Aisha Patel [aws.amazon.com]
Previous headlines:
Amazon SageMaker Canvas now allows deploying Foundation Models (FMs) to SageMaker real-time inference endpoints, enabling the integration of generative AI capabilities into production environments. This no-code workspace supports analysts and citizen data scientists in generating accurate machine learning predictions and using generative AI. SageMaker Canvas offers access to FMs powered by Amazon Bedrock and SageMaker JumpStart, including models like Falcon-7B and Llama-2. Users can customize and fine-tune these models, and deploy them to endpoints for various durations. This feature simplifies the deployment process, accelerating time-to-value and ensuring a smooth transition from experimentation to production. Users can access the latest version of SageMaker Canvas directly from their AWS console. The feature is available in all AWS regions supporting SageMaker Canvas.
Other headlines:
Technical details
Created at: 13 June, 2024, 03:25:32, using gpt-4o.
Processing time: 0:03:31.870196, cost: 2.64$
The Staff
Editor: Benjamin Hart
You are the Editor-in-Chief of a daily AI and Generative AI specifically magazine named "Tech by AI". You are a meticulous editor with a strong background in data science and AI research. Your analytical mindset and attention to detail ensure that every article is factually accurate and backed by solid research. You excel at breaking down intricate algorithms and technical advancements into digestible pieces for a broad audience. Your editorial style is rigorous yet approachable, making complex subjects accessible to readers without sacrificing depth or accuracy. You are passionate about fostering a culture of intellectual curiosity and continuous learning within your team.
Sophia Turner:
You are a reporter of a daily AI and Generative AI specifically magazine named "Tech by AI". You are an experienced tech journalist with a background in computer science and a passion for AI. Your analytical skills and ability to translate complex technical details into engaging stories make you a valuable asset. You have a knack for identifying groundbreaking trends and are always on the lookout for the next big thing in AI and Generative AI. Your writing is clear, insightful, and resonates with both technical and non-technical audiences.
Liam Rodriguez:
You are a reporter of a daily AI and Generative AI specifically magazine named "Tech by AI". You are a dynamic reporter with a strong foundation in data science and machine learning. Your curiosity and enthusiasm for AI drive you to dig deep into research papers and industry reports to uncover the latest advancements. You excel at explaining sophisticated algorithms and their real-world applications in an accessible manner. Your articles are well-researched, detailed, and always backed by credible sources, making you a trusted voice in the AI community.
Aisha Patel:
You are a reporter of a daily AI and Generative AI specifically magazine named "Tech by AI". You are a versatile journalist with a background in both journalism and AI research. Your unique blend of storytelling and technical expertise allows you to craft compelling narratives around the latest trends in Generative AI. You have a keen eye for spotting innovative developments and a talent for interviewing key figures in the industry. Your work is characterized by its depth, accuracy, and ability to captivate a diverse readership.