195 | 🗞️🚨 Huge news from OpenAI + Google

In today’s Brainyacts:

1. GPT-4o is released and FREE! You must watch some of the videos I share below!

2. Google’s Project Astra (crazy cool!)

3. Meta’s AI earbuds and other AI model news

4. ABA list of key generative AI litigation matters and other AI-related content

👋 to new subscribers!

To read previous editions, click here.

Lead Memo

NEW RELEASE: OpenAI GPT-4o (“o” as in omnimodel)

🚨🚨Free users get access to the latest OpenAI model! Go to chatgpt.com now to try it out.

Yesterday (and still today), OpenAI broke the internet. They live-streamed a 15-minute announcement and demo of their new model GPT-4o

What is GPT-4o?

GPT-4o, where the "o" stands for "omni," represents a huge leap forward. The model excels in understanding and generating human-like text, recognizing live images and audio, and responding to multiple prompts.

Key Features of GPT-4o

1. Enhanced Speed and Cost-Effectiveness: GPT-4o is twice as fast (and 50% cheaper for developers) than its predecessors.

2. Improved Text, Vision, and Audio Capabilities: The model provides better responses, recognizing and generating text, images, and audio with greater accuracy and nuance.

3. New Voice Assistant Feature: This feature can read facial expressions, translate in real-time, handle interruptions, and respond to multiple prompts. It enhances the interactive experience, making ChatGPT more like a human assistant.

4. Real-Time Conversational Fluency: GPT-4o understands context and can shift gears during a conversation, handling interruptions smoothly. This makes it feel like a real conversation, not a transaction.

5. Versatile Applications: GPT-4o can understand videos and audio, create 3D animations, and answer complex queries.

Practical Applications of GPT-4o

GPT-4o's features make it suitable for a wide range of applications. Here are some examples:

• Real-Time Translation: GPT-4o can translate languages in real-time, making it a valuable tool for global communication.

• Educational Support: The model can solve math problems, provide explanations, and assist with learning, making it an excellent resource for students and educators.

• Entertainment: THIS IS WORTH WATCHING! So many things happening here. GPT-4o can engage in games like Rock Paper Scissors, adding a fun element to its capabilities.

• Customer Service: Businesses can use GPT-4o for customer support, offering quick and accurate responses to customer inquiries.

• 🚨🚨Interview Prep: You have to listen to this one! Listen to the voice and how it is interacting with inflection, humor, and substance. It see his hair!

Availability and Accessibility

OpenAI is initially launching the new voice assistant feature to a small group of trusted ChatGPT Plus subscribers before a wider rollout. However, most features of ChatGPT Plus are now available to free users, including web browsing, code interpretation, file and image uploads, and memories. Additionally, free users can access GPT-4o, albeit with some limitations (approximately 16 messages per 3 hours compared to 80 messages for paid users). Furthermore, a ChatGPT desktop app for Mac is now available, with a Windows version coming later this year.

Here is everything OpenAI shared yesterday in 2 minutes:

Spotlight

Google Previews Real-time Vision and Audio Capability

Earlier today, Google shared a preview of Project Astra - a new multi-modal AI that processes the context (visual and audio) you are in.

Watch this for a breakdown.

AI Model Notables

► Siri in iOS 18 is rumored to be 'more conversational and versatile' thanks to AI

► Microsoft Places uses AI to find the best time for your next office day for hybrid (virtual and in-office) working

► Google and HP aim to bridge remote worker gap with Project Starline

► Meta's Next Wearable May Be Earbuds With AI And Built-In Cameras

► Claude is now available in Europe:

• Previously available to just US and UK users, European users can now access and interact with Clause across multiple platforms, including a website (claude.ai) and an iOS app

• Claude can reportedly support and understand context in multiple different languages, showing fluency in French, German, Spanish, Italian, and other European languages

• Anthropic has announced paid plans that will unlock access to the advanced Claude 3 Opus model and the Team plan, which is designed for businesses

News You Can Use:

➭ The U.K.'s AI Safety Institute has released a new testing platform to strengthen the safety evaluations of advanced AI systems. The platform aims to provide a comprehensive suite of tools and resources for assessing the safety and reliability of AI models, focusing on areas such as robustness, interpretability, and alignment with human values.

➭ AI systems are learning how to deceive humans intentionally (MIT research paper).

➭ AI washing: A Microsoft VC’s warning about dubious, and rising corporate artificial intelligence claims.

➭ The ABA provides a summary of cases filed related to generative AI – from copyright to privacy to torts.

Who is the author, Josh Kubicki?

Some of you know me. Others do not. Here is a short intro. I am a lawyer, entrepreneur, and teacher. I have transformed legal practices and built multi-million dollar businesses. Not a theorist, I am an applied researcher and former Chief Strategy Officer, recognized by Fast Company and Bloomberg Law for my unique work. Through this newsletter, I offer you pragmatic insights into leveraging AI to inform and improve your daily life in legal services.