AI News Update: OpenAI Spring Event and Google IO Announcements 🎉

OpenAI Spring Event Highlights

• OpenAI Spring Update
  • Released on Monday.
  • Full breakdown video available titled "Chat GPT's Amazing New Models Feel Human and it’s Free."

GPT-40 Announcement

• GPT-40 (Omni)
  • Multimodal model handling text, audio, video, and images.
  • Faster, better model compared to the previous GPT-4 model.
  • Availability:
    • Free for all ChatGPT users with advanced features available (vision models, internet browsing, memory, etc.).
    • ChatGPT Plus Members: Early access to features with five times more outputs.

Developer Features

• API Features
  • GPT-40 API is twice as fast, 50% cheaper, and has a five times higher rate limit.

Conversational & Voice Capabilities

• Her-like Features
  • Demo showcasing conversational abilities resembling Scarlet Johansson’s voice from the movie "Her."
  • Math problem-solving demo using image recognition and showing step-by-step solutions.
  • New desktop app with advanced functionalities (e.g., accessing screen interactions).
  • Ability to change speaking emotions and voices during interaction.
  • Recognizes facial expressions to understand user emotions.

Image Generation & Other Capabilities

• Image Generation: E.g., generating legible text in images (whiteboard).
• Character Consistency: Ability to use same generated character across different scenarios.
• Miscellaneous Features: Photo to caricature, text-to-font, 3D object synthesis, brand placement.

Ilia Sutskever's Departure from OpenAI

• Ilia Sutskever Leaving OpenAI
  • Co-founder, pivotal in Sam Altman’s firing and return had stepped down.
  • Speculation: Discontent with commercialization direction of OpenAI.
  • Left on good terms; planning a meaningful new project.
• Resignations Following Ilia
  • Key members of the super alignment team also resigned.

Google IO Highlights

Gemini 1.5 Release

• Gemini 1.5
  • Two versions: Pro (more intelligent) and Flash (faster).
  • Both versions have a 1 million token context window with plans for 2 million.

Project Astra

• Project Astra Demo
  • Mobile phone identifying objects and remembering visual information.
  • Functions similar to a heads-up display; evident in the glasses shown in demo.

Notebook LM and Image Generation

• Notebook LM: Organizes and explains user’s data in a podcast-style interaction.
• Imagine 3: New text-to-image model with enhanced realism, approaching mid-journey level.
• Veo Video Generation Model: Generating 1080P+ minute-long videos, available soon.

AI Integrations in Google Suite

• Google Search: Enhanced with multi-step reasoning and context-based responses.
• Gmail and Google Meet: AI features for summarizing and organizing email content.
• Ask Photos: Context-aware responses based on photo content (e.g., identifying swimming lessons).
• Scam Detection in Phone Calls: Detects and warns users about possible fraud.

Additional News

Meta Exploring AI Earbuds

• AI-assisted Earphones: Meta working on concept similar to AR glasses but in earbud form.

Microsoft Build Event Upcoming

• Expectations
  • Likely announcements about integration of GPT-40 across Microsoft products.
  • Co-pilot features enhanced with voice and GPT-40 functionalities.

Other Companies

• Anthropic: Hired Instagram co-founder Mike Krieger as Chief Product Officer.
• Hume: Released "Chatter," an interactive podcast experience.

Closing Notes

• Event Season: Multiple upcoming AI events including Microsoft Build, Cisco, Qualcomm, and Apple WWDC.
• Newsletter and Giveaway: Subscribe to stay updated on AI news and win tech gadgets.
• Thank You: Appreciating viewers and subscribers for their engagement.

Stay tuned for more updates from the AI world! 🤖