top of page
Yet Another Agency

AI News #2 - Deep Livecam, Flux AI and Ideogram, Google Gemini Update

Autorenbild: Joschua ZiethenJoschua Ziethen

Aktualisiert: 1. Okt. 2024

Welcome back to the second episode of AI News with Josh! A lot has happened since our last update, and I'm excited to share some of the most intriguing developments in the world of artificial intelligence.



1. Deep Livecam: Real-Time Face Transformation

A Screenvideo showing a person using Deep Livecam to transfer Elon Musk's Face onto their own in realtime
Demo of Deep Livecam. https://github.com/hacksider/Deep-Live-Cam

First up is Deep Livecam, a groundbreaking software that allows you to transfer the face of any image onto yours in real time. Imagine attending video meetings looking like Elon Musk or any other person you choose—all by simply finding a picture online and overlaying it onto your face.


The software is available for everyone on GitHub, complete with detailed installation instructions. It's a fascinating tool to experiment with, so give it a try and see whose face you can wear!


2. Flux AI and Ideogram: Advancements in Text Generation within Images

An AI generated image saying "Ai News by Josh" written into sand on a beach
Generated with Flux AI

Originally, I planned to highlight Flux AI, an impressive image-generating AI model known for its high-quality outputs and exceptional ability to create text within images. However, another model called Ideogram has also emerged, offering similar capabilities.


Both models produce stunning images and have made significant strides in generating legible text—a challenge that has long perplexed AI developers. This advancement is a big deal because it overcomes the common critique that AI-generated images can't handle text well.


You can try both models for free up to a certain limit. Beyond that, you can opt for paid access or integrate them into your projects using their APIs. They're both fun to experiment with, so dive in and explore what you can create!


3. Google's Gemini Update: A Leap Forward in AI Capabilities

A Marketing Banner saying "The Gemini era"
Google Gemini

Gemini Advanced with 1 Million Token Context Window

One of the most remarkable updates is the introduction of Gemini Advanced, featuring a context window of up to 1 million tokens. But why is this such a big deal?


A context window determines how much information the model can "remember" during a conversation. Every word or piece of data uses tokens. With a capacity of 1 million tokens, Gemini can store an enormous amount of information to maintain context in interactions.


For comparison, ChatGPT currently has a context window of up to 128,000 tokens. Exceeding that limit can cause the model to crash or fail to provide accurate responses. With Gemini Advanced, users can have more extensive and complex conversations without losing context. To access this feature, you can upgrade your account to Gemini Advanced.


Device-Built Gemini AI for Android Phones

Google has also introduced a device-integrated version of Gemini AI for Google and Android phones. This is their answer to Apple's recent advancements in AI integration.


With this update, you can:

  • Compose and send emails.

  • Read and manage your Gmail.

  • Access and update your calendar.

  • Seamlessly communicate with other apps.


YouTube Integration

A screenshot of a person asking Gemini "create a list of the foods she eats in this video" in connection with a youtube video
Screenshot from the Live Demo of Gemini + Youtube

A particularly impressive feature is Gemini's ability to connect with YouTube. While watching a video, you can ask the AI questions about the content. For example, "List all the dishes the chef prepared in this video," and Gemini will provide you with an accurate list.


Gemini Live: Real-Time Conversational AI

Google also unveiled Gemini Live, which offers real-time conversational capabilities similar to those introduced by OpenAI earlier this year. This feature allows for live, interactive chats with the AI, complete with selectable voices, enhancing the overall user experience.


These features are currently available on Android devices, with broader support likely coming soon.


Closing Thoughts

That wraps up this episode's AI news highlights. I hope you found these updates as exciting as I did. The world of AI is rapidly evolving, and there's always something new and groundbreaking on the horizon.


Try out these new AI tools and features for yourself, and stay tuned for the next episode of AI News with Josh!


Until next time,

Josh


Taken from LinkedIn Video "AI News Episode 2" , Transcribed by Riverside.fm, written to a blog post by ChatGPT, Human-Proofed by Josh

bottom of page