The fastest way to build AI apps

• Writer Framework: build Python apps with drag-and-drop UI

• API and SDKs to integrate into your codebase

• Intuitive no-code tools for business users

Start building with AI Studio

Today’s top AI Highlights:

1. Figure’s new AI robot; “The world's most advanced AI hardware”

2. OpenAI released structured outputs in OpenAI API with a new GPT-4o model

3. OpenAI lowers DevDay expectations, says no new model will be released

4. Deploy and monitor AI apps with 2 lines of code

& so much more!

Read time: 3 mins

Latest Developments

The AI Robot Designed for the Real World 🦾

The AI robotics company Figure has just revealed their latest creation: Figure 02, a significantly upgraded humanoid robot. This new robot boasts a completely redesigned hardware system focusing on enhanced capabilities and extended operational time. Figure claims this robot is built for the real world and offers improvements in everything from battery life to processing power. You'll want to check out the video - it’s similar to the predecessor but with more refined components.

Key Highlights:

1. Speech-to-Speech Reasoning - Figure 02 can engage in spoken conversations with humans thanks to onboard microphones and speakers connected to custom AI models developed with OpenAI. This is the primary user interface for the robot.

2. Onboard Vision Language Model (VLM) - Equipped with an onboard VLM, Figure 02 can understand and interpret visual information from its six RGB cameras in a way that mimics human common sense reasoning.

3. Onboard GPUs - Figure 02 has 3x the computation & AI inference available on-board compared to Figure 01. This enables real-world AI tasks to be performed fully autonomously.

4. In-house designed Hands - The latest human-scale hands are equipped with 16 degrees of freedom and human-equivalent strength which enables a wide range of human-like tasks

5. Extended Battery Life - A new custom-designed 2.25 KWh battery pack gives Figure 02 a 50% increase in operational time compared to its predecessor. Figure estimates this will allow the robot to work for approximately 20 hours a day, a major step toward practical deployment in various industries.

OpenAI's API Guarantees Structured Data Outputs 📊

OpenAI has just released Structured Outputs in the API that ensures model-generated outputs will exactly match JSON Schemas provided by developers. OpenAI has introduced a new model, gpt-4-02024-08-06, which scores a perfect 100% on complex JSON schema following evals. The model is better than its predecessor (as per the LiveBench Leaderboard) and is 50% cheaper! You can now build more robust applications using the OpenAI API, relying less on workarounds and more on predictable, structured data.

Key Highlights:

1. Increased Reliability - OpenAI utilizes constrained decoding, converting JSON Schemas into context-free grammars to guarantee outputs that precisely match the developer's specifications.

2. Native SDK Support - The Python and Node.js SDKs have been updated to natively handle Structured Outputs, allowing for easy schema definition and response parsing using Pydantic or Zod objects.

3. API Price Reduced - Switching to the new gpt-4o-2024-08-06 model offers significant cost savings: 50% on inputs ($2.50/1M input tokens) and 33% on outputs ($10.00/1M output tokens) compared to gpt-4o-2024-05-13.

4. Implementation Options - Choose between function calling with the strict: true setting or directly specifying the response_format parameter with json_schema to control how the model generates structured outputs.

Quick Bites

1. Amazon Music has a new AI feature called Topics that identifies and tags the topics discussed in a podcast. When you tap on the Topic tags in those podcasts, you’ll also see related podcasts on the same topic. This would help to discover content better.
   

2. Chinese have another text-to-video generation AI app. ByteDance has released Jimeng AI, now available to Chinese users on the App Store. This is after two Chinese companies have already made their text-to-video apps publically available: Kling AI and Ying.
   

3. OpenAI has settled the speculations about the release of a new model. In an update of their highly anticipated DevDay 2024, OpenAI has changed the format of their flagship event to a series of on-the-road events in San Francisco, London, and Singapore.

   • No new model will be released at this DevDay. Instead, the event will focus on their API and dev tools.

   •  The discussions will cover best practices in model customization, evaluations, steerability, scaling, and other topics, led by AI experts.

   • Meet the OpenAI product and engineering teams, see their models in action, and explore innovative projects from top developers and startups.

   Developers who want to attend the event can apply till August 15. If your application is selected, the registration fee is $450.

Tools of the Trade

1. Keywords AI: A unified DevOps platform that simplifies building, deploying, and scaling AI applications with just two lines of code. It offers infrastructure setup, model evaluation, prompt management, performance monitoring, and more.

2. Narrator by Rendernet AI: Create hyper-realistic lip-synced videos to tell your story without appearing on the camera. Just upload the source video, choose the voice, and give it the script. Check out the examples here.

3. assistant-ui: It is a set of React components for creating AI chatbots, supporting various model providers like OpenAI and Google Gemini. It integrates with tools like Langchain, TailwindCSS, and React Hook Form for easy setup and customization.

4. Awesome LLM Apps: Build awesome LLM apps using RAG for interacting with data sources like GitHub, Gmail, PDFs, and YouTube videos through simple texts. These apps will let you retrieve information, engage in chat, and extract insights directly from content on these platforms.

Hot Takes

1. I feel bad for all the people working day and night to produce minor tweaks to LLMs that will soon be obsolete. ~
   Pedro Domingos
   

2. One pair of founders I talked to today were students at a top US university. I asked what percent of students don't use AI to write their papers for them. They said max 20%. Professors have given up trying to forbid it. ~
   Paul Graham

Meme of the Day

OpenAI vs Anthropic

That’s all for today! See you tomorrow with more such AI-filled content.

Real-time AI Updates 🚨

⚡️ Follow me on Twitter @Saboo_Shubham for lightning-fast AI updates and never miss what’s trending!

PS: We curate this AI newsletter every day for FREE, your support is what keeps me going. If you find value in what you read, share it with your friends by clicking the share button below!