MS putting AI everyone in windows 11 with Copilot

interesting post about a few approaches to improve RAG app quality responses

Hi folks. Here is something that you might find it as a useful resource in genAI and LLMs. We're excited to present our latest eBook, "Unlocking the Power of LLMs", co-authored with Greg Coquillo, Product Manager at Amazon. This comprehensive guide offers a deep dive into the intriguing world of Large Language Models (LLMs) and their diverse applications spanning numerous industries. It incorporates valuable insights from industry leaders such as Stripe, Morgan Stanley, Digits, Canva, Elsevier, and Genesis Cloud. Here's what's inside: ✅ Understand the fundamentals and significance of Large Language Models. ✅ Explore the training and fine-tuning processes behind these models. ✅ Examine real-world case studies illustrating the impact of Large Language Models ✅ Gain insights into their role in transforming finance, healthcare, and e-commerce industries. 👉 Want the free ebook? Just head on over to the link provided below, fill out the form and I’ll send the eBook straight to your inbox. Link: https://www.truefoundry.com/ebook-unlocking-the-power-of-llms

Today, U.S. President Biden signed an executive order that places certain guardrails around the development and use of AI technologies. During a White House speech, Biden called AI the "most consequential technology of our time," saying he would do everything possible to "promote and demand responsible innovation." More: • Under the order, any company creating an AI model with possible national security risks must report it to the government and give information on the testing and security measures taken. That sharing of test data only applies to future models such as GPT-5. • In-use models like GPT-4 or Google's Bard are also still subject to other parts of the order, including equity, non-discrimination, consumer protection, and worker-related provisions. • The order will also require federal agencies to assess large AI models before using them and share reports on how they plan to incorporate AI.

Today, OpenAI shared dozens of new additions and improvements, and reduced pricing across many parts of our platform. These include: • New GPT-4 Turbo model that is more capable, cheaper and supports a 128K context window • New Assistants API that makes it easier for developers to build their own assistive AI apps that have goals and can call models and tools • New multimodal capabilities in the platform, including vision, image creation (DALL·E 3), and text-to-speech (TTS) https://openai.com/blog/new-models-and-developer-products-announced-at-devday?utm_source=substack&utm_medium=email OpenAI also announced custom GPTs and an App Store-style GPT Store with a rev share model coming later this month - Link

https://www.theverge.com/2023/11/4/23946353/generative-ai-copyright-training-data-openai-microsoft-google-meta-stabilityai - interesting takes on the copyright laws of data used for AI training. Where do you stand on this issue?

Good comparison of LLM cost https://tomtunguz.com/gm-saas/

OpenAI's GPT-4 and Anthropic's Claude 2 can pass the legal ethics exam, a test nearly all U.S. lawyers are required to pass, according to LegalOn Technologies research. GPT-4 scored the highest out of all large language models tested, achieving a 74% accuracy rate, which is 6% higher than the average (mean) human test-taker. Both GPT-4 and Claude 2 scored above the estimated passing threshold for the MPRE in all states, which ranges from 56% to 64%. https://www.businesswire.com/news/home/20231116912443/en/Generative-AI-Passes-the-Legal-Ethics-Exam-in-Study-by-LegalOn-Technologies

YouTube will begin requiring creators to label videos containing "realistic" AI-generated content. YouTube said users who violate the new rule will be "held accountable," with potential punishments including removing videos or demonetizing creators' channels.

Check out this article about NeevCloud, which has announced the launch of the country's first AI SuperCloud. The company aims to make India self-reliant on AI and Supercomputing and help users in resolving India's core challenges by deploying AI cloud infrastructure including 40,000 GPUs and storage worth $1.5 billion by 2026. Read the full article here: https://hostingjournalist.com/neevcloud-unveils-ai-supercloud-in-india/ ➡️Also, you can reserve your Nvidia H100 GPUs with us! At just $1.69/Hr. Lowest price offered, globally. Tech Specs ✅ 8xH100 SXM ✅ 640GB GPU Memory ✅ 2TB DDR4 RAM ✅ 200 vCPUs ✅ 32TB NVMe Storage ✅ 3200 GBPs InfiniBand You can connect with me anytime. Just drop a text!

The U.S., U.K., and other global powers released non-binding guidelines for how companies should develop and deploy AI models. The international agreement, issued on Sunday, urges companies to make their AI systems "secure by design." More: • 18 countries, including the U.S., U.K., Australia, Canada, France, Germany, and Japan, approved the agreement, while China did not sign. • The 20-page document stresses the importance of designing and using AI systems with security as a top priority. • While not legally binding, the agreement offers general recommendations, including monitoring AI model infrastructure, vetting software suppliers, detecting tampering pre and post-release, and training staff on cybersecurity. • The document doesn't address more controversial AI topics, like image-generating models, deepfakes, or copyright issues. • The guidelines were spearheaded by the U.K. National Cyber Security Center and jointly developed with the U.S. Cybersecurity and Infrastructure Security Agency.

JPMorgan Chase & Co. reports that its AI tools are now generating revenue for the bank. The financial services giant initially set a target of $1B in AI-generated "business value" for 2023, which it later raised to $1.5B. More: • JPMorgan is using AI to give personalized advice to cardholders and share insights with client-coverage teams, among other initiatives, according to Teresa Heitsenrether, the bank's chief data and analytics officer. • It's also been developing a GPT-powered investment adviser. The bank applied for a trademark for "IndexGPT," which would use AI-powered cloud software to analyze and choose financial securities for customers. • Heitsenrether said JPMorgan has "hundreds" of AI use cases now, "some of which are somewhat close to coming to fruition." • The bank is also working closely with U.S. regulators to make sure its generative AI projects are well-controlled.

• By 2030, AI in banking is projected to generate over $300B in global business value, with around $99B in Asia Pacific, $86B in Europe, and $78B in North America.

Lots of cool things happening in the world of AI this week!

Google's Bard chatbot is now being powered by Gemini, the company's most advanced AI model. According to Google, it's the biggest improvement to Bard since its launch in March. CEO Sundar Pichai said the model will eventually come to Google's search engine, ad products, Chrome browser, and beyond.

This is an amazing read to LLM architecture with mind blowing good interactive viz: https://bbycroft.net/llm

They should have been using ML monitoring 😬 https://time.com/6564726/ai-chatbot-dpd-curses-criticizes-company/

Hello Good Noon, Team 👋 Struggling to find the perfect free AI art tools to unleash your creativity? But don’t worry! Check out some excellent free AI art tools that are user-friendly, capable of producing stunning artwork, and best of all, completely free. 👇 https://medium.com/@toolplate.ai/7-best-free-ai-art-generators-you-can-use-right-now-f52e6916b5cb Show your love with Likes, Comments, and Shares!

Hey, does someone have a nice resource on how to create a detailed AI policy for your organization?

Get 35 unrestricted learning credits every month to use on courses and hands-on labs through Google Cloud Skills Boost.The best part: these credits will continue to renew every month, so you can keep learning and earning skill badges https://www.linkedin.com/posts/jose-praveen_googlecloud-developers-practitioners-activity-7205489062344810496-u5I6