Welcome to your daily newsletter on AI

What have we got for you today?

• Grok - A quick overview

• Meta Motivo - Humanoid control

• Tutorial with Microsoft AI Orchestrator

We are 100% free!

And with your support, we create more FREE content!

Please share with a friend, or 10!

🎯 RELEASES 🎯

Groq
A quick overview

Synopsis

Yesterday, We looked at LangChain’s report on The State of AI in 2024  where we say that Groq was listed as #4 in the Top 10 LLM providers. Groq also has its Appgen suite, a tool to build micro apps. So today, let’s look at what Groq has to offer.

Core Observations

The Groq Cloud Developer Console offers documentation, quickstart guides, and an API cookbook with code examples and tutorials to jumpstart your development process

1. Studio or Chat: You can choose to use it as a regular user (like ChatGPT) or dive into its advanced features.

2. API Keys: $25 of free credits until the end of 2024

3. Community: strong community of builders for discussion on advanced usage.

4. Model Compatibility: supports OpenAI models, Llama (Meta, up to 128k context), Gemma (Google, 8k context), Mixtral (32k context), and Whisper (OpenAI, scalable language processing). Also, the “view code” is amazing.

5. Parameters: typical parameters like temperature (0=deterministic 2=creative), max tokens (depends on model), stream (tests a constant stream like API mode), json mode for outputs, and more.

Broader Context

With ChatGPT and Google releasing many features, Groq has followed the same pattern. Groq offers its developer studio, similar to Google which streamlines development for builders. If you use Groq2 on X, it’s 100% free!

Meta
Meta Motivo: Meta FAIR’s Game-Changing AI Model for Complex Humanoid Control

Synopsis

Meta FAIR recently introduced Meta Motivo, a groundbreaking Behavioral Foundation Model designed to control virtual physics-based humanoid agents in executing complex, whole-body tasks. This model is interesting for applications in robotics, animation, and virtual simulations. It holds the potential to redefine how humanoid agents perform dynamic, human-like actions with precision and adaptability.

Core Observations

As a behavioral foundation model, Meta Motivo is designed to generalise across various tasks, offering broad usability in industries like gaming, virtual reality, and robotics.

1. Zero-Shot Control Capability:
   enables humanoid agents to perform complex, full-body tasks without requiring specific task training.

2. Physics-Based Whole-Body Coordination:
   The model excels in controlling humanoid agents grounded in virtual physics, making them capable of realistic and synchronised movements in simulated environments.

3. Open-Source Access:
   Open source and available on GitHub with a Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0) License. This means that you cannot use it for commercial purposes, professional or research.

Broader Context

Meta Motivo addresses one of the most challenging aspects of humanoid robotics translating high-level behaviors into realistic, physics-based actions. We saw earlier this week A Glimpse at the Future of Robotics and AI where Google announced a partnership with a robotics company Apptronik - among other efforts mentioned. Many people have been claiming that AGI (artificial general intelligence) has been reached in 2024, but can chatGPT open a door? Our opinion here is that AGI has not been reached, and until it can interact with the world, open doors like these cats, then we’ve still got a long ways to go. Hey AI, can you catch me a fish? (and the answer is no). So, these efforts - the leading edge of research - aim to bridge that gap.

Read the paper

Try the demo

Trending

• Collier Dolittle prize - half million prize to understand animal languages

• xAI is testing a standalone iOS app for its Grok chatbot

• Human versus autonomous car race ends before it begins

We are 100% free!

And with your support, we create more FREE content!

Please share us with a friend!

⚙️ BUILDERS BYTES ⚙️

Microsoft AI Orchestrator

What will you learn today?

Learn how to build AI-powered agents for autonomous cloud management using Microsoft's AI Ops Lab framework. We'll explore how these agents optimise operations, automate decision-making, and improve reliability in cloud systems.

Key Takeaways

1. Task Based: APIs are a set of documented tools, e.g., get logs, get metrics, and exec shell, designed to help the agent solve a task

2. Scale-Up: orchestrator has privileged access to the deployment and can take arbitrary actions (e.g., scale-up, redeploy) using appropriate tools (e.g., helm, kubectl) to resolve problems on behalf of the agent

3. Control: orchestrator calls workload and fault generators to create service disruptions, which serve as live benchmark problems.

from aiopslab import Orchestrator
class Agent:
    def __init__(self, prob, instructs, apis):
        self.prompt = self.set_prompt(prob, instructs, apis)
        self.llm = GPT4()
 
    async def get_action(self, state: str) -> str:
        return self.llm.generate(self.prompt + state)
 
#initialize the orchestrator
orch = Orchestrator()
pid = "misconfig_app_hotel_res-mitigation-1"
prob_desc, instructs, apis = orch.init_problem(pid)
 
#register and evaluate the agent
agent = Agent(prob_desc, instructs, apis)
orch.register_agent(agent, name="myAgent")
asyncio.run(orch.start_problem(max_steps=10))

To view the full code, to here

Trending

• An AI App that Pay parking tickets - great look at AI Agents

• Autoregressive Video Generation without Vector Quantization

• Normalization techniques for LLM- Combining Pre-LN and Post-LN (code, paper)

🤩 COMMUNITY 🤩

Tutorials

• CrewAI + Groq tutorial

• Groq API Crashcourse for beginners

Talks & Events

• TedTalk (with audio only) - AI pioneer Mustafa Suleyman shares his vision for the future of AI

• Google DeepMind - Project Astra: Exploring a Universal AI Assistant with Greg Wayne

THANK YOU

Found something cool?
Want something different?

Our Mission at AlphaWise

AlphaWise strives to cultivate a vibrant and informed community of AI enthusiasts, developers, and researchers. Our goal is to share valuable insights into AI, academic research, and software that brings it to life. We focus on bringing you the most relevant content, from groundbreaking research and technical articles to expert opinions to curated community resources. 

Looking to connect with us?

We actively seek to get involved in community with events, talks, and activities. Email us at [email protected] 

Looking to promote your company, product, service, or event?