Elon debuts Grok-2 and Grok-2 mini to supercharge X

In today’s newsletter:

• xAI Releases Grok-2: Key Features, Comparisons and Impact

• Tutorial: How to produce high quality prompts with LangGPT

• Google releases Pixel 9 phone series with AI-powered Android update

• Huawei testing a Nvidia chip competitor, Ascend 910C

• 4 AI tools in the spotlight:

  • Cisco Meraki Cameras with AI-Powered Object Detection

Read time: 4-5 mins

xAI unveiled Grok-2 and Grok-2 Mini today, its latest AI models designed to enhance user experiences on the X platform with real-time information access and advanced capabilities like image reasoning and generation. This release marks a significant step in integrating AI with social media, offering premium subscribers exclusive access to advanced features.

This release marks a significant step in integrating AI with social media to revolutionize how users interact with content. For example, it enhances search capabilities, allows users to gain deeper insights into posts, and improves reply functions.

Read more about its key features, Comparisons and Impact → here.

The ability to make good use of existing large models relies heavily on high-quality prompts. However, writing high-quality prompts is still a craft at this stage and relies too much on personal experience.

With the release of the GPT-4 model, the reliance on prompts has been reduced. At the same time, its more powerful basic capabilities provide a good foundation for writing more powerful prompts. High-quality prompts are becoming more powerful and more complex.

So is it possible to master some basic rules and concepts, and some programming patterns (similar to object-oriented programming) just like learning programming, so that you can efficiently write prompts with good performance and stability?

The answer is yes.

To find out how to write reusable prompts, check out this tutorial → here.

Revolutionizing Surveillance: Cameras with AI-Powered Object Detection

In today's fast-paced world, simply having surveillance cameras isn't enough. What if your cameras could not only record but also understand and interpret what they see?

This isn't a vision of the future—it's happening now. Below 11 lines of python code (e.g. run in Raspberry-pi) shows how simple it is to use IP camera’s RTSP stream to detect objects using COCO data set.

To see the full tutorial by Felix Lam, Head of Sales Engineering at Cisco Meraki (APAC) → here.

News from Asia:

• Huawei readies new AI chip to challenge Nvidia in China as its latest processor, Ascend 910C, is currently being tested Chinese telcos and internet companies.

• Samsung announces $7 billion profit, largely attributable to rising demand of high-end chips and AI chip prices.

• China pushes for favourable, orderly development of GenAI as reports shows more than 600 million now use LLMs amid rapid GenAI adoption.

• Sakana AI, a Japanese AI startup, has released The AI Scientist which runs a simulated review process for evaluation, mimicking the human scientific community's peer-review process.

  • You can read the paper → here

News from the West:

• Google joins the AI-enhanced hardware club with Pixel 9 phone series and AI-powered Android update announcements.

• deepset launches Studio for Architecting LLM Applications with Native Integrations to deepset Cloud and NVIDIA AI Enterprise.

• What could have been: Recent reports revealed that Intel had the chance to acquire a 15% stake in OpenAI for $1 billion, with the potential for an additional 15% stake in exchange for producing hardware at cost.

• NVIDIA's new Blackwell GB200 AI servers affected by 'component shortage', leading to expected short supply in Q4 2024.

• AMD completes all-cash $665 million Silo AI acquisition.

• Payzen, an AI startup tackling healthcare affordability in the U.S., closed a $232 million Series B.

Trending Tools & Apps

1. Smart Cameras by Cisco Meraki helps you stay ahead by staying on top of your security system with unmatched visibility and control via AI-powered object detection.

2. PaddleBoat sharpens sales pitches by simulating cold-calls with realistic AI buyers. Choose from a list of AI buyers or create your own with customizable personalities, objections, and instructions tailored to your business context.

3. Frontend AI allows you to ask for a component or upload an image, and instantly receive ready-to-use code without lifting a finger.

4. Flowith is the AI for deep work. Surpassing traditional chat-based tools, it streamlines tasks on a multi-thread interface powered by a most advanced agent framework. The intuitive canvas and smart framework boost productivity, helping users stay in the flow.