This Week in AI: NotebookLM upgrades, new Pikaffects, and Google’s Imagen 3
All the AI news for creatives
Good day, humans.
In today’s Newsletter:
1️⃣ You can now add prompts to NotebookLM
2️⃣ Highlights from Adobe Max 2024
3️⃣ Google launches Imagen 3
4️⃣ Pika Labs introduces new effects
5️⃣ AI Video of the Week: The Fall of Gondolin
6️⃣ Steal this Midjourney style
Let’s go.
You can now customize your NotebookLM podcast
It’s official: one of this year’s biggest AI sensations isn’t about image generation or chatbots—it’s a study tool. NotebookLM’s audio summary features have captivated users by transforming dry documents and media into lifelike, podcast-style discussions. Now, Alphabet’s surprise hit returns with major improvements.
You can now tell the AI hosts who to talk to and what to focus on (e.g., generate a discussion aimed at teens, or focus on just one part of a larger document). You can also listen to audio summaries while multitasking in NotebookLM.
Best uses for NotebookLM:
People have been using the app for these amazing purposes:
Upload a document on a complex topic, and the hosts will explain it to you.
Upload your LinkedIn profile, and the hosts will analyze your career.
Upload your diary, and the hosts will give you their feedback.
Learn more: Turn your articles into engaging podcasts with this free Google tool.
Adobe Max 2024 Highlights
This week at Adobe Max 2024, Adobe introduced over 20 new features in its Creative Cloud suite. Here’s a quick summary of the key announcements:
Firefly Video Model: Adobe’s first publicly available video model designed to be commercially safe, as it has been trained only on content they have the rights to use.
Generative Extend: Adobe Firefly’s AI video model has been integrated into Premiere Pro with a feature called Generative Extend, now in beta. This tool allows you to add up to two extra seconds of video footage and up to 10 seconds of audio to your clips.
Distraction Removal: Photoshop’s new feature can automatically detect and remove unwanted objects, like people, cables, and wires, with a single click. No manual selection required.
Objects on Path: Adobe Illustrator now lets you easily attach, arrange, and move objects along any shape, giving you more control and flexibility in layout design. The mockup tool for visualizing designs on 3D models is now available to all users, making it easier to see how your designs will look in real life.
Google launches Imagen 3
Google DeepMind has unveiled Imagen 3, the latest iteration of its AI image generation model. The model can now produce images with a level of photorealism that is almost indistinguishable from human-created works. However, it cannot generate images of specific or identifiable people.
Pika Labs introduces new effects
Pika Labs has added 4 new effects to its previous lineup: it now includes the ability to crumble, dissolve, deflate, and make video subjects disappear with a "ta-da" effect, essentially making a subject vanish behind a cloth. Users can simply upload an image to the site, and Pika 1.5 will turn it into a video with a corresponding animation.
Learn more: Pika Labs is back in the AI video game with these new effects.
AI Video of the Week: The Fall of Gondolin
Steal this Midjourney style
To use it, add --sref 782391 at the end of your prompt in Midjourney. Your images will be generated with this style.