• VasuGPT
  • Posts
  • The Biggest Week In AI⚡️

The Biggest Week In AI⚡️

With the surprise release of GPT-4o and the huge announcements coming out of Google I/O 2024, This is the biggest single week till now! 🔥

Welcome! It’s Friday, May 17th.

Do you know?🤔OpenAI's president, Greg Brockman, has shared the first public image generated using the new GPT-4o model.The image shared by Brockman is a significant improvement in quality, photorealism, and accuracy of text generation compared to OpenAI's previous DALL-E 3 model.

On that note, Here is today’s menu:

  1. AI Spotlight: What's Going On With OpenAI? 🤯

  2. Tool Talk: Text-to-Speech Arena 🪄

  3. Prompting Curiosity: The AI Prompt of the Week 📝

  4. AI Resources: Grow Every Day by 1% 📚

Major AI Announcements from Google I/O 2024

Marc Rebillet kicking off Google’s I/O event in a rainbow-colored bath robe

Why it Matters?Google’s I/O 2024 Developer’s Conference kicked off this week, and while OpenAI's GPT-4o announcement took some of the wind out of their sails, Google still came out swinging with a huge volume of AI announcements.

Much of this conference is aimed at developers, so we've filtered all the uber-technical news and we're just telling you everything that's happened at Google I/O you'll actually care about:

  • Gemini models received significant upgrades. The new Gemini 1.5 Pro boasts a massive 2 million token context window, along with enhanced performance in code, logic, and image understanding.

  • For those using Gemini Advanced, there’s an exciting new feature that allows the creation of personalized AI personas called ‘Gems’ from simple text descriptions, similar to OpenAI's GPTs.

  • Google introduced Veo, a new video generation model capable of producing over 60-second, 1080p videos from text, image, and video prompts. This puts it in direct competition with OpenAI’s Sora. This Veo model powers the new VideoFX text-to-video tool, which supports storyboard creation and music addition and is launching in a private preview in the U.S.

  • The new Imagen 3 text-to-image model also made its debut, offering improved detail, text generation, and natural language understanding. ImageFX, now powered by Imagen 3, is available via a waitlist.

  • Google also showcased its new AI agent project, Project Astra. This real-time AI agent prototype can see, hear, and take actions on a user’s behalf. It was demonstrated with a voice assistant responding to visual and auditory inputs, showcasing advanced reasoning and recall capabilities. Public access for Astra is expected through the Gemini app later this year.

  • Google Search is also getting a major upgrade, featuring expanded AI Overviews, advanced planning capabilities, and multi-step reasoning to break down complex questions and speed up research.

It should be noted that most of these are "coming soon" and are not available for consumers just yet. But for those heavily invested in the Google ecosystem, it's time to celebrate! These are huge improvements to Google's current AI offerings.

Speaking of Google's AI offerings--we noticed that Google has by far the largest catalog of AI tools, but there's nowhere you can go to see "everything AI" that Google has to offer.

So instead of waiting around for Google to make it, we created a fantastic resource compiling everything Google AI-related and organising it beautifully for you here in the free area of The AI Advantage Community.

ChatGPT-4o 🪄

Use Case: Omnimodal AI Companion

Why you should care?In their Spring Update Livestream, OpenAI announced the release of the latest addition to the ChatGPT family: GPT-4o. It's faster and cheaper than GPT-4 Turbo, and it combines text, audio, and visual inputs and outputs into a single "omnimodal" model. (That's where the 'o' comes from!)

During the stream, OpenAI employees showed off GPT-4o's upgraded Voice Mode, shocking and impressing the entire world. The live interactions and demos from OpenAI's blog went viral immediately, with many online comparing GPT-4o to the AI chatbot Samantha from the 2013 film 'Her'.

Perhaps the most surprising of all was OpenAI's announcement that GPT-4o would be made freely available to the public, though not all at once. Here's the current status of the GPT-4o rollout:

  • GPT-4o (for text) available to most paid users

  • GPT-4o (for text) free version started rolling out yesterday

  • Image Generation today is still DALL-E 3

  • Image Upload is the new GPT-4o

  • Improved Web Browsing and Code interpreter are available today

  • GPTs still use GPT4 although a few select people report an update to GPT-4o with a brand new creation interface

  • Voice input remains the old Whisper

  • Voice output remains the old tts-1

  • Mac App downloadable for many but unusable for now

  • Phone app not available yet

There's no official roadmap on the GPT-4o rollout from OpenAI, so nobody's sure exactly when these features will be available.

Due to this, expect more GPT-4o coverage coming soon! We'll do a full breakdown of the effectiveness of each tool as they come out, whether that's in this newsletter, on our YouTube channel​

Prompt of the Week: Highlight covers for Instagram 📝

Topic- Highlight covers for Instagram.

Benefit- Instagram highlight covers are the solution to this problem. They're like the cover of a book for your Stories, giving your profile a polished and professional look. By using minimalist icons, brand colours, or even photographs related to your content, you can make a strong first impression. These covers not only help reinforce your brand identity but also organise your content, making it easier for followers to browse through what interests them.

I need Instagram highlight cover ideas for [Your Topic/Keyword]. Any suggestions?

Output

Ensure you give ChatGPT detailed information about your specialty, intended audience, and brand for more customized and relevant recommendations.

Ensure you give ChatGPT detailed information about your specialty, intended audience, and brand for more customised and relevant recommendations.

AI Resources 📚

AI program aims to break barriers for female students (link)

Must-read sci-fi books about AI to fill your summer reading list (link)

Scientists use generative AI to answer complex questions in physics (link)

CEO Sal Khan on why he thinks AI can become every student's personal tutor (link)

NVIDIA, Teradyne and Siemens gather in the ‘City of Robotics’ to discuss autonomous machines and AI (link)

P.S. You can still get The BEST ChatGPT Prompting Cheat Sheet For Beginners 2024 (no email opt-in).

That's all for this week folks…

If you have a second, I’d appreciate it if you could rate this email from Cool to Not-so-cool. Just poll and let me know!

Stay Curious, Hustlers!

See you next Friday.Much Love, Vasu

Reply

or to participate.