Headlines You Should Know
Congratulations, You’re Now a Director
First, AI was just generating the written word. Then images. AI audio is still trending upward. Now, video is the next frontier of AI content creation coming soon to a computer near you. This morning, Google announced a new app called Google Vids, an AI-powered video creation app for work. That last phrase, “for work,” is doing a lot of heavy lifting. Google’s demos show this as a way to create sales training videos or marketing material by leveraging Google Docs, Sheets, or Slides as the backbone of the content.
Unlike Sora, OpenAI’s text-to-video AI model that looks like it’s more about entertainment, Google Vids is geared toward enterprise use of AI. It’s interesting to note that Google is making it easy for Vids to interact with the assets you already have instead of forcing users to create ChatGPT-style prompts. Presumably, this will make it easier to create the content you’re aiming for whenever Google makes it widely available.
Video tools like Google Vids will be the next arrow in the quiver for communication pros tasked with creating content. The lessons learned from using other AI tools and platforms will be helpful in mastering a new powerful tool to create engaging video content with the efficiency and adaptability they have come to expect from AI in text, image, and audio generation.
Elsewhere …
- Gaming Finds New Role as AI’s Open Ethics Lab
- LISTEN: The AI Compliance Check Every Financial Services Marketer Needs
- AI Use for Hiring, Firing Employees ‘Problematic,’ NYU Professor Says
- Google Partners with Bayer on New AI Product for Radiologists
- Arm CEO Warns AI Could Consume 25% of U.S. Power by 2030
- George Carlin AI-Enabled Fake Comedy Taken Down in Settlement
Tips and Tricks
You can finally tweak DALL·E images in ChatGPT
What’s happening: OpenAI quietly added a feature that can save a lot of time when trying to get the right image: selective editing of DALL·E outputs.
Why it’s important: Sometimes (most times?) DALL·E doesn’t quite capture everything you had in mind for an image. If you prompt it again, it starts from scratch. The new feature allows you to pinpoint certain elements to change, like removing or adding objects from the scene, fixing a typo, or changing a color.
Try this: After your initial prompt generates an image, click on it to bring it into full view. You’ll now notice that there’s an icon in the upper right corner of a paintbrush making a swirl. That’s the “select” tool that allows you to choose an area of the image to edit. Once you’ve selected the area, type in the box what you’d like to change, and DALL·E will edit only within the selected area.
Quote of the Week
“When TV first arrived, media companies used it to broadcast radio programs or pre-recorded theatrical plays. Watching a play on TV isn’t a great way to use the medium, and eventually TV developed its own conventions of close ups, quick cuts, and scene changes.
“The same mistake is being made with AI right now. It is being used by some misguided publishers to generate static articles — using a new medium to mimic a legacy medium. This mistake is made again and again throughout history, and the key to success is to explore relentlessly until you find things that are only possible in the new medium and then lean into them to create magic.”
— Jonah Peretti, Founder and CEO at BuzzFeed, in his annual letter to shareholders