_{Welcome back for the Latest AI Drops!}

Today’s Drops:

• Google unveils MedLM, AI models for health care

• Google launches AI Studio For Developers & Imagen 2 with text and logo generation

• Quality 3D Object Generation from Single Images

• Trending on X

• Trending Startup “Harvey”

• AI Headlines: Axel Springer collaborates with OpenAI to enhance journalism using AI, Google Cloud partners with Mistral AI for hosting language models, Google launches Duet AI with Gemini model for coding, and Glean, leveraging OpenAI's models, nears a $2 billion valuation in advanced data search.

• Trending GitHub Projects

Read Time: 4.5 minutes

Google unveils MedLM, AI models for health care

Google has unveiled MedLM, a new suite of AI models tailored for healthcare. This development reflects Google's effort to capitalize on the healthcare AI market, competing with Amazon and Microsoft. MedLM,

Built on the Med-PaLM 2 model, it includes both large and medium-sized models designed for different healthcare tasks.

• The larger model is intended for complex tasks requiring in-depth knowledge and computational power

• The medium-sized model optimized for more agile, real-time functions like summarizing doctor-patient interactions.

This suite reflects Google's approach to offering a range of AI tools rather than a single "one-size-fits-all" solution, catering to various needs within the healthcare sector.

Companies like HCA Healthcare have already started implementing MedLM to enhance their operations. HCA utilizes MedLM to assist emergency medicine physicians in documenting patient interactions and converting transcriptions into structured ER provider notes. This application has shown promise in reducing the time doctors spend on paperwork, potentially improving efficiency in healthcare delivery.

The World Health Organization has expressed concerns about the rapid adoption of AI in healthcare, citing the potential for harm due to incorrect information, disinformation, and data privacy issues. Despite these challenges, Google maintains a commitment to cautious and responsible deployment of these technologies, aiming to support healthcare professionals while ensuring broad access to the benefits of AI in healthcare.

Source

Google launches “AI Studio” For Developers & “Imagen 2” for text and logo generation

1. AI Studio Launch: Google has introduced AI Studio, a web-based tool for developers, evolving from its previous incarnation as MakerSuite. This platform is built on the Gemini model and offers robust prompt and chatbot development features. AI Studio stands out with its free tier, scalable to paid options, and is part of Google's extensive AI ecosystem. It supports Gemini Pro and Gemini Pro Vision models, boasting flexible settings for model output, safety, and diverse workflows for prompt creation. This tool is designed to be accessible to developers of all skill levels and integrates seamlessly with other Google AI offerings like Vertex AI.

2. Imagen 2 Release: Enhancing its AI-driven image generation capabilities, Google has released Imagen 2 to select Google Cloud customers. Developed in collaboration with Google DeepMind, Imagen 2 introduces superior image quality and new functionalities, including multilingual text and logo rendering. It also has advanced training techniques for comprehending extended prompts and incorporates invisible watermarks for added security.

Source & Source

Quality 3D Object Generation from Single Images

Stability AI has introduced Stable Zero123, a model designed for 3D object generation from single images. This model enhances the quality of 3D understanding and view generation, surpassing its predecessor, Zero123-XL. Key innovations include an improved training dataset, elevation conditioning for better predictions, and a new data loader for increased training efficiency. Stable Zero123 is based on Stable Diffusion 1.5 and is available for non-commercial and research use. It offers a significant speed-up in training compared to Zero123-XL and supports the creation of textured 3D meshes.

Check out the demo video

Trending on X

Trending Startup “Harvey”

Harvey, a legal AI startup, is currently negotiating a funding round of $70-80 million, which would value the company at $700 million including the new investment. This valuation marks a significant increase from its previous valuation announced in April. The funding round is being co-led by venture firm Kleiner Perkins and Elad Gil, a solo investor who has previously backed Harvey in its seed and Series A rounds. Existing investors Sequoia Capital and OpenAI Startup Fund are also participating.

Source

More AI News

Axel Springer collaborates with OpenAI to integrate advanced AI into journalism, leveraging their media brands to enrich AI-generated content and foster a sustainable future for independent journalism. Link

Google Cloud has formed a partnership with Mistral AI to host and scale the startup's large language models on its advanced AI infrastructure, leveraging Google's robust security and privacy capabilities. Link

Google's Duet AI for Developers, now generally available and soon to integrate the Gemini model, is an AI-assisted coding tool offering advanced code completion and generation in over 20 languages, set to compete with GitHub's Copilot and enhanced by partnerships with 25 companies; it's free until January 2024, then $19 per user per month annually. Link

Glean, a startup by ex-Google engineers using OpenAI's models, enhances corporate internal searches and is nearing a $2 billion valuation with potential new funding, reflecting significant growth and investor interest in its advanced data search solutions.Link

Welcome to the innovative world of GitHub projects!

DeepEval is an open-source evaluation framework designed for language model applications akin to Pytest. It supports various models like RAG and Llama2, offering a wide range of local evaluation metrics such as hallucination, answer relevancy, and bias. DeepEval enables users to easily create custom metrics, evaluate large datasets efficiently with minimal code, and integrate with tools like Confident AI for continuous assessment. Its key features include automatic integration with existing systems, efficient bulk dataset evaluation, and tools for continuous improvement and debugging of LLM applications.

Colossal-AI making large AI models cheaper, faster, and more accessible

AWS CLI (Command Line Interface) is a unified tool to manage Amazon Web Services directly from the command line. It allows users to control various AWS services, execute scripts on these services, and automate tasks.

PandasAI is a Python library that adds Generative AI capabilities to pandas, the popular data analysis and manipulation tool. 

Jetson-containers offers machine learning containers optimized for NVIDIA Jetson and JetPack-L4T. It includes support for ML frameworks like PyTorch and TensorFlow and is designed for easy customization and building of containers. This resource is particularly useful for AI/ML development on NVIDIA Jetson platforms.