- The Rundown AI
- Posts
- Elon Musk's xAI reveals the newest chatbot
Elon Musk's xAI reveals the newest chatbot
PLUS: Microsoft's AI set to run locally on PCs
Rowan Cheung
March 29, 2024
Welcome, AI enthusiasts.
Elon Musk’s AI startup xAI just revealed the latest version of its Grok chatbot, which will soon be rolling out to users on the X platform.
If benchmarks hold up and a potential partnership with Midjourney pans out — the chatbot could be the newest competitor to ChatGPT and GPT-4. Let’s dig in…
In today’s AI rundown:
xAI and Elon Musk unviel Grok 1.5
Microsoft Copilot to run locally on future PCs
Generate slides for your presentations using AI
AI21 Labs unveils open-sourced Jamba
7 new AI tools & 4 new AI jobs
More AI & tech news
Read time: 4 minutes
LATEST DEVELOPMENTS
XAI
Image source: xAI
The Rundown: xAI just announced Grok-1.5, the latest iteration of its open-source large language model, boasting improved reasoning capabilities and a massive 128,000 token context length.
The details:
Grok-1.5 has notable improvements in coding and math, achieving high scores on benchmarks like MATH (50.6%), GSM8K (90%), and HumanEval (74.1%).
The model will soon be available to early testers and existing Grok users on the X (formerly Twitter).
Rumors suggest that Grok-1.5 may include an "Analysis" button, which can summarize entire threads and replies on X.
Why it matters: If benchmarks are any indication of LLM performance, Grok-1.5 should be nearly as capable as GPT-4. But Grok's true potential lies in its ability to leverage the real-time data on X, and if Grok’s recent talks with Midjourney pan out, Grok 1.5 could be a serious competitor in the AI chatbot frenzy.
TOGETHER WITH SECTION
The Rundown: Learn how to harness the power of AI to streamline workflows and get more done in less time.
In this free workshop, AI expert Ed Ortega will guide you through:
Setting up a custom "AI workspace" for optimal efficiency
Identifying tasks where AI excels, saving time and effort
Leveraging AI for first drafts, brainstorming, and more
Integrating AI into your daily routine
Enroll now to start maximizing your productivity with AI.
MICROSOFT COPILOT
Image source: Microsoft
The Rundown: Intel has confirmed that Microsoft's Copilot AI will soon operate locally on PCs, reducing reliance on cloud processing and potentially boosting performance and privacy.
The details:
Future AI PCs will need NPUs (Neural Processing Units) with at least 40 TOPS (Tera Operations per Second) to handle more elements of Copilot locally.
Microsoft prioritizes running Copilot on the NPU instead of the GPU to minimize the impact on battery life.
The trend of local AI processing is also extending to smartphones, with Google's Pixel 8 also featuring the Tensor G3 chip for on-device AI.
Why it matters: The shift towards local AI processing could enhance performance, privacy, and cybersecurity, setting the stage for a potential "TOPS war" among chipmakers competing to meet the growing demand for powerful, efficient AI hardware.
AI TRAINING
The Rundown: In this tutorial, you will learn how to use Gemini in Google Slides so it generates slides, images, and text for you. The best part? It’s free.
Step-by-step:
Visit Google Slides. You will need to sign up or log in with your Gmail account.
Select “Blank Presentation” to create one from scratch.
Click on the Gemini icon at the top right. Describe the slide, hit Enter, and your slide will come to life.
Click “Insert” if you are happy with it or regenerate it.
That’s it! To modify or add an image, click the Gemini icon again and ask for an image with a description.
PRESENTED BY BRILLIANT
The Rundown: Brilliant's bite-sized, interactive learning makes it easy to level up in minutes a day—whenever, wherever.
Brilliant is your shortcut to:
Building skills in AI
Thinking like a data scientist
Learning programming, technology, and more
Want an edge in the future? Start sharpening your mind now. Join 10 million people around the world and start your 30-day free trial today.
AI21LABS
Image source: AI21 Labs
The Rundown: AI21 Labs just introduced Jamba, an open-source AI model that merges the Mamba Structured State Space (SSM) architecture with components of traditional transformer architecture, creating a powerful hybrid system.
The details:
Jamba boasts an impressive 256K context window, equivalent to around 105,000 words or 210 pages.
The model delivers 3x faster processing on long contexts compared to transformer-based models of similar size.
Jamba is highly efficient, requiring just a single GPU with 80GB of memory to run — despite its large context window.
Jamba’s weights are available on HuggingFace here.
Why it matters: Jamba's innovative approach makes it the only model of its size capable of handling 140K context on a single GPU, opening up new possibilities for businesses and researchers looking to leverage state-of-the-art LLMs without the need for extensive computing resources.
NEW TOOLS & JOBS
🛍️ Vetted AI- Free AI shopping assistant to find the best products on the web at the best prices*
🎨 Weavely- Build and design web-based forms in Figma
🤖 AI Chatbot by Zapier- Create an intelligent AI-powered chatbot for free
⛏️ Pickaxe- No-code platform for building AI apps to embed into your website
🚀 Airtable AI- Integrate AI models into your apps and workflows
📣 AYAY AI- Digital marketing assistant automation
🎯 mentor- AI-powered goal management for personal growth
🎯 Mistral AI - Talent Acquisition
⚙️ Fiddler AI - Staff Backend Engineer
📜 OpenAI - AI Policy Lead & Program Manager
🤝 Synthesia - Account Executive (Mid-Market)
*Sponsored listing
QUICK HITS
Google.org (the nonprofit arm of Google) just launched a $20 million funding program for non-profit organizations developing generative AI technologies.
Amazon plans to spend almost $150 billion in the next 15 years on data centers to handle the expected explosion in demand for AI apps.
Researchers in London are developing an AI tool to sort dogs into 5 categories in a bid to match puppies with their perfect owners.
New York City will soon start testing technology that uses AI to detect firearms at subway turnstiles.
All US federal agencies are now required to have a Chief AI Officer and AI governance boards to oversee AI systems and ensure responsible use as part of the new Office of Management and Budget guidance.
Microsoft is building new safety features into Azure AI Studio to make it harder for people to trick AI chatbots, including "prompt shields" to detect and block deliberate attempts to make models behave in unintended ways.
THAT’S A WRAP
SPONSOR US
Get your product in front of over 500k+ AI enthusiasts
Our newsletter is read by thousands of tech professionals, investors, engineers, managers, and business owners around the world. Get in touch today.
FEEDBACK
How would you rate today's newsletter?Vote below to help us improve the newsletter for you. |
If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.
Reply