Google AI reintroduces human image generation after historical accuracy outcry

Pivot 5: 5 stories. 5 minutes a day. 5 days a week.

1. Google AI reintroduces human image generation after historical accuracy outcry

Google's Gemini AI model is resuming the generation of images of humans after being paused in February due to concerns over historically inaccurate racial depictions. The new Imagen 3 model, which was first announced in May, will roll out the generation of images of people to Gemini Advanced, Business, and Enterprise users in the coming days.

A version of the model, complete with human image-generation capabilities, was recently made available to the public via the Gemini Labs test environment without a paid subscription. Google does not support the generation of photorealistic, identifiable individuals, depictions of minors, or excessively gory, violent, or sexual scenes. In quick tests, the new Imagen 3 system avoided many of the "historically inaccurate" racial pitfalls that led to Google's pause in its generation of human images.

Read the full story here

2. Google’s custom AI chatbots have arrived

Google is allowing Gemini subscribers to create custom chatbots called Gems, which can serve various roles such as gym buddy, cooking partner, and writing editor. Users can give Gems distinct personalities and specialties by describing a set of instructions.

Premade Gems are available for users who don't want to create a custom chatbot. Google is rolling out Gems to Gemini Advanced, Gemini Business, and Gemini Enterprise users in over 150 countries and languages.

Read the full story here

3. Apple’s new iOS developer beta lets you remove objects from pictures using AI

Apple has introduced a new AI feature called Clean Up, allowing users to remove objects from photos without affecting the rest of the image. The feature uses smart detection and can be drawn, brushed over, or manipulated using a circle.

It also generates the background and removes shadows or reflections. Apple Intelligence features have been in development since July, and are only available to users in the U.S.

Read the full story here

4. Midjourney teases hardware product in a new form factor

Midjourney has announced its plans to enter the hardware market. The company, founded by David Holz, has hired Ahmad Abbas, who previously worked at Leap Motion and Apple. Midjourney's official X account has posted multiple tweets revealing its plans for the device, which is not a pendant, and that it has "multiple efforts in flight" for more details. 

The company has also mentioned opportunities for more form factors. Midjourney and Holz have been teasing and trolling about the project on X, with no timeline for the announcement. The company is just one of many companies exploring the future of AI-driven hardware.

Read the full story here

5. Generative AI Transformed English Homework. Math Is Next

OpenAI's generative AI chatbot, ChatGPT, has impacted classrooms and changed how teachers approach writing homework. Students are now using AI more often to complete math homework. The Gauth app, owned by ByteDance, has millions of downloads and is primarily focused on mathematics.

Students can point their smartphone at a homework problem, and the AI model generates a step-by-step guide with the correct answer. While the app struggles with higher levels of math, researchers are working to improve AI's abilities in this sector.

Read the full story here

Advertise with Pivot 5 to reach influential minds & elevate your brand

Get your brand in front of 80,000+ businesses and professionals who rely on Pivot 5 for daily AI updates. Book future ad spots here.