📰 News 📰 

Unveiling the World's Most Powerful Open Source AI Model

Databricks introduces DBRX, surpassing Meta's Llama 2 as the most potent open source large language model. The team spent $10 million and months training DBRX, outperforming in various benchmarks and even challenging OpenAI's GPT-4. By open-sourcing DBRX, Databricks aims to promote transparency and innovation in the AI field, contrasting with the secretive approach of other tech giants. The model's release signifies a significant advancement in AI development and may lead to further research and understanding of AI capabilities.

Read More At: Wired

Claude 3 Overtakes GPT-4 in Chatbot Arena

For the first time, Anthropic's Claude 3 has surpassed OpenAI's GPT-4 in Chatbot Arena, marking a significant milestone in the evolution of AI language models. This achievement was highlighted by the excitement on social media, underscoring the competitive advancements in the AI industry. With Claude 3's rise to the top and the introduction of smaller models like Haiku, the AI community is seeing a shift in leadership to a diversity of vendors, fostering innovation and competition. This development not only challenges OpenAI's dominance but also signals a vibrant future for AI technologies, as Chatbot Arena continues to be a critical platform for comparing and understanding the nuances of AI model performance.

Read More At: Ars Technica

Accelerating AI: Oriole Networks' Breakthrough in LLM Training

Oriole Networks, a University College London (UCL) spinout, has secured £10M in seed funding to revolutionize the training of Large Language Models (LLMs) with their innovative AI "super brains." By employing light to connect thousands of AI chips, they've created a network that operates as a singular, powerful machine, enabling LLM training to be 100 times faster and significantly more energy-efficient. This advancement not only promises to reduce AI's hefty energy demand but also to speed up processes like algorithmic trading, presenting a major leap forward for AI and data centre industries.

Read More At: The Next Web

Amazon's $4 Billion Investment in Anthropic Concludes

Amazon has finalized its $4 billion investment in AI startup Anthropic, establishing itself as the primary cloud provider for the company. This strategic move allows Anthropic to utilize Amazon Web Services (AWS) Trainium and Inferentia chips for its machine learning models. Unlike Microsoft's investment in OpenAI, Amazon's contribution is a direct cash investment without cloud credits. This investment is part of a broader trend among tech giants to fuel AI development, aiming to attract and retain customers on their cloud platforms. Anthropic, known for its generative AI model Claude and a focus on building safer systems, has been a significant player in the AI space alongside competitors like OpenAI.

Read More At: The Register

🧠 Research 🧠 

BioMedLM: A 2.7B Parameter Language Model Trained On Biomedical Text

BioMedLM explores whether smaller, domain-specific models can rival larger, general-purpose models in biomedical NLP tasks. Trained on PubMed abstracts and articles, this 2.7 billion parameter GPT-style model demonstrates competitive performance in biomedical question-answering, outperforming or matching much larger models across various benchmarks (e.g., MedMCQA, MedQA, MMLU, PubMedQA, and BioASQ). Notably, BioMedLM scored 57.3% on MedMCQA (dev) and 69.0% on MMLU Medical Genetics. It addresses the drawbacks of large models by being more economical, privacy-preserving, and environmentally friendly, offering a promising solution for biomedicine-focused NLP applications. The paper underscores the potential of medium-sized, domain-focused models like BioMedLM, suggesting a balanced approach towards leveraging domain-specific training data. Future research could explore enhancing BioMedLM's performance on tasks where domain specificity plays a crucial role, further refining its utility for biomedical applications. The model is publicly available on Hugging Face Hub for broader community engagement and development.

Read the paper on arXiv

🔨 AI Powered Tools 🔨 

A run down of the latest SaaS products and services which leverage AI to help you take back some time.

• Bardeen - Automation powerhouse to simplify work processes and execute mundane task without you lifting a finger. Use Bardeen for meetings, marketing, product development, sales & prospecting, recruiting, data research, and personal productivity.

• Hexomatic - Powerful no-code platform designed for web scraping and workflow automation, catering to sales, marketing, and research needs. It leverages AI, including native ChatGPT and Google Gemini automations, to enhance its capabilities.

• Recut - Automatic video editor that significantly reduces editing time by automatically removing silences from videos. Aimed at content creators and educators, Recut is valuable for those producing videos, podcasts, or any audio-visual content.

🗞️ Other News Rollup 🗞️ 

AI Collective Evolution - AI units share knowledge like Star Trek's Borg for rapid learning and transformative technologies.

Saturn's Moon Exploration - NASA's snake robot, EELS, to explore life on Enceladus, a water-rich moon. Innovative design for tight spaces.

AI in Healthcare - Viome uses AI for personalized health solutions, including dietary recommendations and disease detection.

Future of Humanoids: Insights - Humanoid robots evolving with Nvidia's showcases, challenges in deployment, and pilot programs with major companies.

AI Adoption Risks - Enterprise AI adoption up 600%, 18.5% transactions blocked. Zscaler report emphasizes cybersecurity strategy importance.

Startup Success with AI - 86% of founders benefit from AI, reshaping strategies. Startups adopt AI for growth, facing challenges but remaining optimistic.

Creative AI Experiments - Google Arts & Culture launches Food Mood for recipes and Musical Canvas for drawing soundtracks, fostering creativity.

Hopper Leads Generative AI - NVIDIA Hopper excels in generative AI inference, achieving record-breaking performance with H200 GPUs and TensorRT-LLM software.

Data-Driven Transportation -cyvl.ai uses sensors and AI to enhance transportation infrastructure data, partnering with firms for success and growth.

AI Farming Innovation - Orchard Robotics uses AI to collect precise apple crop data, securing $3.2 million for development and testing.

Dark Web Startup - StealthMole secures $7M funding to expand and specialize in Asia-related cyber threats.

🎶 Prompts 🎶 

Prompts used to generate some of this issues images. Unless otherwise stated all images are using Dall-E and ChatGPT Plus.

Thank You for Subscribing

Enjoying what you’re reading? Help me get better so I can continue to provide you with the most relevant content.