• Work With AI
  • Posts
  • Microsoft Starts Beef With Elon Musk and Stability AI Joins the LLM Battle

Microsoft Starts Beef With Elon Musk and Stability AI Joins the LLM Battle

April 20, 2023

⚡ Today’s Highlights

  • 📰 News: Microsoft is getting into beef with Elon Musk and Stability AI is joining the LLM battle

  • 💰 Funding: Coro, EdgeQ, Lexion, Trullion, and Groundlight

  • 🦾 Tools: Can you tell the difference between a person and an AI? Put yourself to the test!

  • 📅 Events: The Stable Diffusion AI Hackathon and the AI in Finance Summit New York end today!

📰 Today's Top Stories

(4 min read) (Source: TechCrunch)

TLDR: Microsoft is dropping Twitter from its advertising platform next week, after Twitter announced that it will begin charging a minimum of $42,000 per month to users of its API, including enterprises and research institutions.

  • Microsoft's decision means that users will no longer be able to access their Twitter account, or create, schedule or otherwise manage tweets through Microsoft’s free social media management service.

  • Twitter's new pricing scheme has faced criticism for pricing out smaller businesses and research institutions. This move is part of Twitter's effort to monetize its platform and boost its revenue streams.

  • In response today, Elon Musk threatened to take legal action against Microsoft for its decision, claiming that Microsoft trained its powerful AI models on illegally acquired Twitter data through its licensing arrangement with the AI outfit OpenAI.

(5 min read) (Source: Stability AI)

TLDR: Stability AI launches StableLM suite of language models, marking its entry into the language model game. It is available in alpha, featuring 3 billion and 7 billion parameter models, both trained on 800 billion data tokens, with larger 15-billion to 65-billion parameter models to follow.

  • The StableLM suite builds on Stability AI's work on image models, including Stable Diffusion, which offered an open-source alternative to proprietary generative AI image models such as DALL-E and Midjourney.

  • The StableLM suite also includes a set of research models that are instruction fine-tuned, using a combination of five recent open-source datasets for conversational agents.

The Big Picture: Stability AI's release of the StableLM suite of language models is a significant step towards making foundational AI technology accessible to all and offers an open-source alternative to proprietary generative AI language models dominated by tech heavyweights such as OpenAI, Meta and Stanford. The StableLM suite has numerous applications, including generating text and code, and could have a major impact on the future of AI design.

(3 min read) (Source: Reuters)

TLDR: Snap has announced that its AI chatbot, My AI, will now be able to respond to users' messages with fully AI-generated images.

  • The addition of AI-generated images to My AI is part of Snap's strategy to deepen its work in AI technology and speed up the development of augmented reality features for Snapchat. By using AI to recommend lenses that can add effects to photos and videos or suggest real-world places to visit on the app's map feature, My AI could help users explore more parts of the Snapchat app.

  • Snap has added safety guardrails to My AI, including restricting a user's access if they repeatedly ask inappropriate or harmful questions and ensuring age-appropriate conversations.

(4 min read) (Source: TechCrunch)

TLDR: Salesforce is working on a generative AI-driven workflow system that can create code based on written requests.

  • The company is developing two new tools, Einstein GPT for Flow and Data Cloud for Flow, which combine generative AI textual interaction and data-driven event triggers to automate workflows with little or no code.

  • The combination of Salesforce's Data Cloud, Einstein GPT, and Flow tools enables users to accomplish complex multi-system workflows from data.

💰 Funding Alerts

🦾 Trending Tools

  • Spectate is an AI-powered all-in-one platform for effortless monitoring, alerting, incident management, and status pages. With insanely fast notifications, phone call and SMS alerts, and global monitoring options, Spectate is the perfect solution for teams who want to focus on resolving issues quickly.

  • Cognosys brings AI-powered agents to your fingertips with a web-based version of AutoGPT. You can now effortlessly create copy output that can be downloaded and enhanced with markdown formatting.

  • My AskAI lets you easily create a customizable ChatGPT chatbot from your own content, without coding. Use it for customer support, research, site/product search, and more. Available as a Zap, Bubble plugin, Slackbot, Teams app, and WordPress plugin. Customize fonts, colors, answer styling, and more.

  • Introducing Regex.ai, the AI-powered regular expression generator that's easy to use, efficient, and entirely free. With an updated interface and helpful features, users can create regular expressions that match specific patterns effortlessly. Gain insight into regex generation reasoning and learn regex basics with in-depth documentation.

  • Can you tell if someone is human or not? Test your ability to distinguish between humans and state-of-the-art AI bots like GPT-4 and Jurassic-2 with this free online game. Chat for two minutes and guess if your conversation partner was a bot or a human. Explore the nature of AI and contribute your feedback to the community.

👀 More Reading

🌎 Fresh Content

1. In this episode of In Machines We Trust, host Jennifer Strong sits down with University of Oxford and Article19 Human Rights Researcher Mahsa Alimardani to discuss “smart cities” and what they might mean for privacy. Mahsa is studying how this is being applied in Iran and goes in-depth on one of the nation’s top smart cities, to learn how its efforts there have evolved over time. (Spotify)

When AI watches the streets

Listen to this episode from In Machines We Trust on Spotify. The term ‘smart city’ paints a picture of a tech-enabled oasis—powered by sensors of all kinds. But we’re starting to recognize what all these tools might mean for privacy. In this episode, we meet a researcher studying how this is being applied in Iran and visit one of the nation’s top smart cities, to learn how its efforts there have evolved over time. We Meet: University of Oxford and Article19 Human Rights Researcher Mahsa Alimardani City of Las Vegas Chief Innovation Officer Michael Sherwood City of Hope Director of Campus Support Operations Mark Reed Sounds: How will artificial intelligence change the cities we live in? - BBC Ideas via YouTube  https://www.youtube.com/watch?v=UXxyCBimRyM  ‘Smart’ cities promise economic and environmental benefits to the developing world - CBC News via YouTube https://www.youtube.com/watch?v=u08A7yiTmu4  Singapore is building a city in China - CNBC via YouTube  https://www.youtube.com/watch?v=iP11XeIV1ZA  Global Smart Cities - The China Current via YouTube  https://www.youtube.com/watch?v=-qmiqHWD6Uc  Footage appears to show Iranian riot police confronting students at university in Tehran - The Guardian via YouTube  https://www.youtube.com/watch?v=BgQshPJohmg  China: facial recognition and state control - The Economist via YouTube  https://www.youtube.com/watch?v=lH2gMNrUuEY  Facial recognition: Concerns over China's widespread surveillance https://www.youtube.com/watch?v=CT6KEy_QXvM Credits: This episode was reported and produced by Jennifer Strong and Anthony Green with help from Emma Cillekens. It was edited by Mat Honan, and mixed by Garret Lang, with original music by Garret Lang and Jacob Gorski. Artwork by Stephanie Arnett.

2. Spotify is the most successful music streaming platform out there and it’s all because of how good it is at predicting peoples’ music tastes. This video from the Wall Street Journal goes into detail about how Spotify’s amazing AI algorithm works and how it puts together its personalized playlists. (YouTube)

4. The Practical AI podcast invited Rajiv Shah, ML engineer at Hugging Face, to discuss how Large Language Models have been evolving as well as several intriguing topics such as in-context learning, reasoning, LLM options, and related tooling. (Changelog)

📚 Research:

 📅 Upcoming Events

  1. The Stable Diffusion AI Hackathon starts today (April 14 - 21, 2023), a 7-day hackathon in which participants must develop a product that leverages the power of Generative AI to solve real-world problems. To participate in the hackathon, you must create the solution using Stability.AI API and Vercel software. 1st place wins $5000 USD, 5000 API tokens, and a virtual coffee chat with @hardmaru!

  2. Check out the AI in Finance Summit New York (April 20-21, 2023, etc.venues - New York, NY) and discover unique insights and technical use cases from AI specialists and data scientists across Banking, Financial Services, and Insurance sectors.

  3. Semantic Layer Summit 2023 (April 26th, 2023, 10:30 AM - 05:00 PM EST / Virtual) is a free one-day virtual event featuring 50+ Enterprise Data Leaders & Top Industry Technologists. Hear advice and best practices from top data leaders, AI & BI technologists, & industry experts in Enterprise data analytics.

  4. EmTech Digital (May 2-3, 2023 / In person on the MIT campus or Virtual), MIT Technology Review’s signature AI conference, provides CIOs, CPOs, and IT decision makers with the technology insight and business expertise to implement AI systems that enhance products, improve customer experiences, and ensure market success. This year’s conference focuses on the game-changing power of generative AI and how it will impact creators and clients alike. We’ll explore the models behind the technology and the legal implications of generated content. We meet leading researchers and explore hands-on case studies of amazing applications of all forms of AI.

🧠 Resources

Being a modern business professional means juggling several different skills to get to the top of your game. It can become very overwhelming. But with AI, you can focus on what you do best while leaving the rest to technology. Our goal is to make difficult tasks easier and more efficient for you. Don't worry if you're not an 'AI Whisperer,' give these prompts a try and let us know what you think. We're excited to hear your feedback! Remember, these prompts are just a starting point. Tailor them to fit your needs and carefully review the outputs to ensure their accuracy and effectiveness. Let's get started!

IT Architect

  • In this role, you will act as an IT Architect and be responsible for integrating a new application or digital product into an existing IT landscape. This task will involve analyzing business requirements, performing a gap analysis, and creating a solution design, physical network blueprint, and deployment environment blueprint. 

  • Prompt: "Your role is to act as an IT Architect and integrate a new application or digital product into an existing IT landscape. You will need to analyze business requirements, perform a gap analysis, and create a solution design, physical network blueprint, and deployment environment blueprint. Please confirm that you're ready to start by saying 'Ready to integrate the new application.'"

Persuasive Email

  • Your task is to create a persuasive email that provides valuable and relevant information to your ideal customer persona about a specific subject, persuading them to take a desired action. The email should have a clear and compelling message. 

  • Prompt: "Your task is to create a persuasive email that provides valuable and relevant information to your ideal customer persona about a specific subject and persuades them to take a desired action. The email should have a clear and compelling message. Please confirm that you're ready to start by saying 'Ready to craft a persuasive email.'"

Customer Development Process

  • Utilize the Customer Development Process framework to create a marketing campaign outline that identifies and validates customer needs for a specific product or service. The outline should describe how you would build and test prototypes to meet those needs and outline the steps you would take to iterate based on customer feedback. Specific tactics and metrics to measure success should be included. 

  • Prompt: "Using the Customer Development Process framework, your task is to create a marketing campaign outline that identifies and validates customer needs for a specific product or service. The outline should describe how you would build and test prototypes to meet those needs and outline the steps you would take to iterate based on customer feedback. Specific tactics and metrics to measure success should be included. Please confirm that you're ready to start by saying 'Ready to create a marketing campaign outline.'"

That's all for today's email! If you want more please follow us at the social channels linked below, or check out our website!

How'd you like today's email?

Login or Subscribe to participate in polls.

Share our newsletter: If you like our work please share/forward this email with your friends, colleagues, and family. It's the best way to support us!

If this email was forwarded to you please sign up here to continue receiving them.

Want your content, product, jobs, or event featured in our newsletter? Reply to this email with the details, and our team will reach out to you.

Do you use AI for work? Tell us how, and you could be featured in our newsletter!

Check out our website for more resources, including a list of AI investors, products, events, and twitter follows. 

For an archive of all our posts, click here.

We'd love to hear from you! You can always leave us comments or feedback by replying to this email!

Powered by AI. Curated and edited by Humans.