ISUX “June” Industry Design Trend Express

In June, the design circle was collectively opened: Google M3 Expressive reshaped the button with “physical bounce”, Airbnb’s 3D Lava icon made the UI breathe, and Apple Liquid Glass unified the visual language of the five major platforms in one go; Midjourney video model V1 is online, clipping “Little Skylark AI” is a one-click hit, Search Live turns search into conversation, and Imagen4 directly makes Gemini an “AI photo studio” that can be taken immediately.

[1] Google releases the M3 Expressive design language

[2] Google updates iOS app icons

[3] Airbnb visually refreshed and launched a new icon animation solution

[4] Byte publishes image Agent “Little Skylark AI” to create a one-click explosive creation artifact

[5] Apple award 2025 was announced, and the national app won the award

[6] Midjourney launches its first video generation model, V1

[7] Google Search Live voice search function is launched, revolutionizing the voice conversation search experience

[8] Imagen4 lands on Gemini, AI image generation enters a new era

[9] Apple unveiled a new style of Liquid Glass, unifying the interface across five major platforms for the first time

How can product managers do a good job in B-end digitalization?
All walks of life have taken advantage of the ride-hail of digital transformation and achieved the rapid development of the industry. Since B-end products are products that provide services for enterprises, how should enterprises ride the digital ride?

View details >

1. Google releases the M3 Expressive design language

3. Airbnb visually refreshes and launches new icon animation solutions 3. Airbnb Visuals Refresh and Launches New Icon Animation Solution3. Airbnb visually refreshed and launched a new icon animation solution

▲Picture from: Material Design

The upgraded core of M3 aims to balance brand personalization with system consistency, providing a “more flexible, smarter, and more inclusive” design system.

Flexible: Break the limitations of style and empower brand expression.
Intelligent: Dynamic color and adaptive layout reduce multi-terminal adaptation costs.
Inclusivity: Enhance accessibility support, ensuring personalization doesn’t sacrifice usability.

▲Picture from: Material Design

To achieve this, the team has made a number of improvements, adding a physics-based motion mechanic that makes buttons and elements look like natural jumps in interactions and transitions, and also enriching a range of shape animations and fonts and color schemes, and the team is also advocating for a more dynamic use of these design elements, and developers can quickly implement new features through the updated Material Design 3 component library and Figma toolchain.

2. Google updates iOS app icons

Google released a new “Gradient G” logo at the I/O developer conference in May 2025, which has been gradually applied to its core products. The gradient design is regarded as a visual symbol of Google’s AI strategy (such as Gemini 2.5 Pro, Project Mariner), embedding design genes for virtual and real integration interactive scenes, and the new logo aims to strengthen cross-platform consistency while conveying the futuristic sense of technological aesthetics through fluid texture.

▲Picture from: App Store

3. Airbnb visually refreshes and launches a new icon animation solution

Airbnb redesigned its app in May 2025 and introduced a brand new icon media format internally called “Lava.” is Airbnb’s new 3D animated icon media format, which is essentially a lightweight video with an alpha channel (transparent background) optimized for UI icons. It was built to overcome the limitations of existing solutions like Lottie (2D vector animation) and traditional video formats. Lava icons support full 32-bit color and transparency, so icons can have soft shadows, glows, and other effects that blend in with the UI. Airbnb has filed a patent specifically for this, and Airbnb’s vice president of design said Lava’s animated icons will both please users and maintain smooth performance. By controlling the format and playback engine, Airbnb ensures that every animated icon looks great and performs reliably.

4. Byte publishes image Agent “Little Skylark AI” to create a one-click hit creation artifact

Compared with similar competitors, Little Skylark’s competitiveness lies in the integration of fragmented tool chains, rather than single-point function optimization, positioning itself as a “content execution agent”, covering the whole process of creation needs such as video generation, digital voice broadcasting, image design, and background replacement.

▲Picture and video from: App Treasure

5. Apple award 2025 winners were announced, and the national app won the award

The list of winners of this Apple award 2025 is the first time that apps from the country have won awards, and there are two, namely Cap words and Taobao. Products that truly impress people have emotional value, design aesthetics and immersive experience.

CapWords comes up with a simple yet revolutionary idea: learn the names of things by taking pictures. With just one tap and fun animations, CapWords can turn everyday objects – coffee cups, traffic cones, cupcakes – into interactive stickers. CapWords reshapes the language learning experience with “real scene collections”, which is an elegant practice of “interest-driven learning” in the AI era.

▲Picture from: Apple |cap words | Taobao design

Taobao offers stunning 3D models in the Vision Pro app that rival physical goods, each boasting smooth interactions, transitions, and intuitive controls. For each 3D product, Taobao takes into account its placement, control, size and function, and provides a super practical side-by-side comparison function.

▲Picture from: Apple |cap words | Taobao design

6. Midjourney launches its first video generation model, V1

▲Picture from: Midjourney official website

Midjourney recently officially released its first “Image-to-Video” model V1 for the public, marking a significant leap from static images to dynamic videos.

▲Picture from: Midjourney official website

When using it, users can select an existing image on the web page and click the “Animate” button to generate a 5-second video clip with one click. Supports “Auto Mode” for quick animations, or switches to “Manual Mode” to customize the direction of motion and lens shift, and flexibly choose between “High/Low Motion” settings. Support video extension up to 21 seconds.

▲Picture from: Midjourney official website

In terms of payment mode, the basic subscription ($10/month) enjoys V1 video access, and the cost per second of video is about equal to the generation of an image (about 8 times the cost of image generation), and premium subscriptions (such as Pro or Mega) can also use “relax mode” to experience unlimited video generation.

▲Picture from: Midjourney official website

This release not only brings AI image creation to life, but also signifies that Midjourney is moving towards its long-term goal of achieving “real-time open-world simulations”.

7. Google Search Live voice search function is launched, revolutionizing the voice conversation search experience

▲Picture from: Google’s official website

Google has officially launched Search Live, the first real-time voice interaction experience with Gemini support. When users enable AI mode in Google apps for iOS or Android, they can simply tap the Live icon under the search box to have a natural conversation with the search engine using their voice. This new feature not only supports instant voice response, but also continuously understands the context and realizes a conversational search link, significantly improving the efficiency and intelligence of information retrieval on the mobile side.

When you’re on the go or busy packing, simply open the Google app, tap the new Live voice icon, and ask questions in a speaking way, such as: “How do I keep my linen dress from wrinkling in my suitcase?” The system will use AI to generate a clear voice answer and support you to ask, “What if it’s already wrinkled?” and other follow-up questions. All relevant information is displayed on the screen at the same time, with easy-to-click web links to help you learn more.

8. Imagen4 Launches Gemini, Entering a New Era in AI Image Generation

Google recently updated its Gemini platform with full access to the latest image generation model, Imagen 4, making it a conversational AI with “instant gallery” capabilities. Users can enter a natural language description (such as “panorama of snowy mountains at sunset” or “vintage-style poster”) into the chat window to get high-quality images in seconds. This means that there is no need to jump to the interface, and the entire process of creation is directly embedded in the conversation flow, greatly improving the convenience and efficiency of use.

▲Picture from: Google’s official website

The generation quality has been significantly improved, and the details and text rendering are more brilliant

The advantages of Imagen 4 are particularly significant: the ability of images to express complex fabric textures, water droplet morphology, animal hair and other details is significantly enhanced; Its rendering of text content is also clearer, making it suitable for design scenarios such as posters and greeting cards that require intuitive text. In addition, the model supports up to 2K resolution output, meeting the needs of digital display and high-quality printing, while generating various styles such as realistic photography, watercolor, cartoon illustration and even abstract art according to prompts.

▲Picture from: Google’s official website

Dialogue generation + real-time adjustment, smarter interaction

This integration gives Gemini Chat conversations the ability to generate and edit images. Users can further modify the generated content during chat, such as adjusting colors, adding elements, or changing styles, allowing for a smooth “chat-and-tune” creative process with just a simple text prompt. Social media feedback shows that this new operating experience is popular with designers and content creators.

▲Picture from: Google’s official website

Gemini（Imagen 4）VS ChatGPT‑4o

When compared to other mainstream image generation tools, Imagen 4 shows clear advantages in realism, text clarity, and speed. However, in terms of generating portraits of specific people and highly customized style transfer, there are still comments that it is slightly inferior to ChatGPT-4o. However, Gemini is gradually building its market advantage with its conversational interaction format, free user support, and cost-effectiveness of 2K output.

▲Picture from: Google’s official website

9. Apple unveils a new style of Liquid Glass, unifying the interface across 5 major platforms for the first time