KIMP’s Picks June 20th: Weekly Highlights In Design, Marketing, AI, & More

Ever find yourself scrolling through social media on a Friday, feeling like you’ve missed out on everything? From the latest AI breakthroughs everyone’s talking about to those crucial social media algorithm changes and all the share-worthy posts. It can feel overwhelming to catch up after a busy week. We get it! But don’t let FOMO get the best of you! KIMP’s Picks is here to save your valuable time. This is our weekly social media recap, designed to give you the week’s essential highlights in just a few quick minutes.

Ready for the KIMP’s Picks June 20th Edition? Let’s dive in!

In the Marketing Realm 

Google discusses crucial small business stats 

In a recent post, Google shared some useful statistics that small business owners need to know. To begin with, Google warns that small businesses are more susceptible to cyberattacks and hence recommends using the right Google tools to keep their information secure, including Google Password Manager. 

The next stat is about the extensive work hours for small business owners and how tools like Google Meet can help boost their efficiency. Finally, Google also reiterates the need for meeting online shoppers where they are. 

Google introduces “Audio Overviews” Search 

Search Engine Roundtable regularly discusses search-related updates. In a recent post, they explored some changes coming to Search. 

Reportedly, Google is launching a new experimental feature called “Audio Overviews” within its Search Labs, aiming to transform traditional search results into concise, conversational audio summaries. This opt-in feature, powered by Google’s latest Gemini AI models, generates short audio responses to certain search queries, similar to its NotebookLM technology.

Users who opt into this Search Labs experiment will see an option on the search results page to generate an audio overview. These overviews provide a quick, hands-free way to grasp information from top search results, with links provided within the audio player for those who wish to delve deeper.

Dove’s campaign wins big discussing authenticity in the AI age 

ADWEEK shared a post celebrating some award-winning campaigns, one of which was Dove’s recent Real Beauty campaign. 

Dove’s long-running “Real Beauty” campaign has evolved for the AI era, earning the Media Grand Prix at the Cannes Lions Festival. In a bold move against the growing dominance of AI-generated imagery, the brand partnered with Pinterest to encourage users to define beauty in their own terms, resulting in algorithm shifts that favored natural, unedited images over digitally created ones.

The campaign garnered significant global attention with 4 billion earned impressions and 500 million views. This campaign highlights the impact of leading with integrity and creating socially aware campaigns that resonate with the target audience. 

An engaging campaign idea from Oreo 

The next post we’re discussing is also an inspirational campaign, and this one is from Oreo. Ads of the World shared this campaign highlighting the power of creativity in advertising. 

In an effort to reconnect with younger audiences across Latin America, without using traditional media or a hefty budget, Oreo found inspiration in an unlikely place: Audio Notes! 

The visual waveform of a voice note, with its dark lines and white gaps, bore a striking resemblance to an Oreo cookie. Leveraging this visual similarity, Oreo encouraged users to send their “most boring” voice notes via WhatsApp. In exchange, participants received discounts on Oreo cookies. 

The campaign from Oreo proves that even the most mundane things, like Audio Notes, can be leveraged into creative campaigns when you know your audience well. 

Cinnamon Toast Crunch targets a young audience with dark humor 

Humor is a universal emotion, and several brands tap into it to add spice to their ads. However, dark humor is not an area many brands venture into as it does not appeal to a broad audience segment. 

Ad Age, in a recent post, highlighted one such brand that capitalized on dark humor successfully to re-engage its Gen Z audience. The post discusses the “Must Cinnadust” campaign from General Mills’ brand Cinnamon Toast Crunch. 

The campaign revives the brand’s early 2000s cannibalistic cereal ads, but cranks up the intensity with serial killer-inspired characters and chilling scenes. 

General Mills embraced the “shock factor,” seeing it as a necessary step to connect with an older Gen Z audience that grew up with the cereal. 

In the Content Marketing Realm 

Semrush discusses local SEO 

When we talk about SEO, we talk about it in a broad sense, but Semrush recently shared a post highlighting how local, national, and global SEO can be different and the nuances that differentiate them. 

It talks about the right strategies for each and the applications along with some tips for brands that wish to blend them. So, if you are looking to boost your website traffic by enhancing your SEO strategies, this is a great read. 

In the Design Realm 

Canva elevates the AI video generation game with “Create a Video Clip” 

Canva has launched “Create a Video Clip,” a new AI-powered feature that allows users to generate cinematic-quality video clips with synchronized sound from simple text prompts. This tool, integrated directly into Canva’s homepage, is powered by Google’s latest video generation model, Veo 3.

“Create a Video Clip” seamlessly fits into Canva AI’s existing conversational workflow, allowing video generation to be part of the same creative ecosystem as text, images, and designs. 

Users can instantly generate 8-second video clips that feature cinematic visuals and smooth motion. Each generated clip automatically opens in Canva’s Video Editor, providing tools for customization such as Brand Kit integration, text overlays, and music additions.

Canva to acquire MagicBrief for AI-powered creative intelligence 

Canva has announced its intent to acquire MagicBrief, an Australian creative intelligence startup founded in 2022. This acquisition aims to enhance Canva’s offerings by integrating AI-powered performance data and optimization capabilities directly into its platform to create a more seamless workflow for content creation.

By bringing MagicBrief’s technology into the Canva Visual Suite, Canva aims to close the loop between content ideation and its impact, allowing users to not only create engaging, on-brand content but also analyze and optimize it using real-time performance data.

This will take Canva closer to other holistic creative ecosystems like Adobe’s suite of tools. 

A brand identity based on emotional design 

Paris-based micro-roastery Datura launched a striking new brand identity ahead of its debut at the Paris Café Festival, created in collaboration with French design studio Brand Brothers. The goal was to craft a look that blends sophistication with an alternative edge, avoiding the familiar cues of micro-roastery branding while preserving the allure of the name.

They explored the physical structure of the lowercase “datura”, noting four circular letters and three vertical separations. This rhythmic pattern creates a tension between refinement and brutalism, highlighted by the elegant interweaving of specific letter pairs. 

This brand identity design project carries valuable lessons in embracing abstract storytelling and designing for clarity and emotional connection. 

In the AI Realm 

Google launches “Search Live” with voice mode 

Google has launched Search Live with voice input within the Google app for Android and iOS. Available to users in the U.S. enrolled in the AI Mode experiment in Labs, this new feature allows users to have free-flowing, back-and-forth voice conversations with Search, enabling exploration of links from across the web.

Search Live is designed for convenience, particularly for users who are on the go or multitasking. By tapping a new “Live” icon in the Google app, users can verbally ask questions and receive AI-generated audio responses. 

This move continues Google’s push to make search more natural, ambient, and multimodal. 

The Gemini 2.5 models are now generally available 

Google has announced a significant expansion of its Gemini 2.5 model family, making Gemini 2.5 Flash and Gemini 2.5 Pro officially stable and generally available across various platforms. This includes Google AI Studio, Vertex AI, and the Gemini App.

In addition to this, Google is launching a preview of Gemini 2.5 Flash-Lite, touted as its most cost-efficient and fastest 2.5 model to date. This new “Lite” version boasts higher quality across coding, math, science, reasoning, and multimodal benchmarks compared to its predecessor, 2.0 Flash-Lite.

According to Google, this new model is well-suited for high-volume, latency-sensitive tasks due to its lower latency than most previous models. 

Midjourney announces first video AI model 

While Midjourney made waves in the AI image generation segment several months ago, the company has not made any big moves recently. Breaking that streak, it has announced V1, its first-ever video model. 

Priced at $10/month, the new feature lets users animate images using simple prompts, making the experience fun, intuitive, and widely accessible.

Dubbed “Image-to-Video,” the tool allows both auto-generated and manual animations, with settings for high or low motion. Each job produces four 5-second videos, which users can extend up to 20 seconds. Even images uploaded from outside Midjourney can now be animated, giving creators more flexibility.

OpenAI rolls out record mode in ChatGPT 

ChatGPT is rolling out a new “Record mode” feature to its Pro, Enterprise, and Edu users, currently available on the macOS desktop app. This functionality enables ChatGPT to transcribe and summarize audio recordings, such as meetings, brainstorming sessions, or voice notes.

Once an audio recording is completed, the system generates a summary, which is saved as an editable “canvas” within the chat history. These canvases can then be transformed into various useful outputs, including project plans, emails, or even code. Additionally, ChatGPT can reference past recordings and their transcripts to provide more contextually relevant responses in subsequent conversations. 

OpenAI enhances ChatGPT’s Canvas feature with a download option 

OpenAI has rolled out new capabilities for Canvas ChatGPT. Users can now download content from Canvas in multiple file types, including PDF, Microsoft Word (.docx), and Markdown (.md). For code generated in Canvas, users can directly export it in relevant programming file types such as .py for Python, .js for JavaScript, and .sql for SQL. Canvas is accessible across ChatGPT’s web, desktop, and mobile applications.

Higgsfield launches “Canvas” for precise AI image editing 

Higgsfield has introduced Higgsfield Canvas, a new state-of-the-art AI-powered image editing model designed to give users pixel-perfect control over product placement, editing, and layout. This tool aims to drastically reduce the time spent on image editing tasks that usually take time. 

Canvas allows users to directly “paint” products onto an image with high precision. After uploading a picture, users can select an area and size it to their liking for seamless product integration. It also supports additional image editing features like swapping clothes, tweaking details, and even changing the faces of subjects. 

Big announcements made at MiniMax Week 

MiniMax AI’s MiniMax Week event has unveiled exciting updates. Firstly, MiniMax-M1, their new LLM, boasts the world’s longest context window with 1 million input tokens and 80,000 output tokens. 

MiniMax also introduced MiniMax Agent, a general intelligent agent built for complex tasks. Reportedly, the agent has been in internal use for 60 days, becoming a daily tool for over 50% of their team.

Finally, MiniMax officially launched Hailuo 02, its new video generation model. This model offers native 1080p resolution, state-of-the-art instruction following, and extreme physics mastery, making it capable of generating highly intricate scenarios like gymnastics.

Proactor unveils AI Agent 

AI company Proactor has introduced what it terms the “world’s first proactive AI agent,” designed to operate with context awareness, memory augmentation, and real-time action, all before a user explicitly asks. This AI agent doesn’t wait for prompts but actively participates in conversations, offering immediate transcription and live summaries.

The core capability of Proactor lies in its ability to instantly analyze discussions to identify potential needs and actionable tasks, whether they are directly stated or merely implied. It can proactively research topics as they arise and execute identified tasks within the flow of a conversation. 

Hunyuan 3D 2.1 launches

Hunyuan has released 3D 2.1, positioning it as the world’s first fully open-source, production-ready 3D generative model capable of producing physically based rendering (PBR) assets. 

The model enables cinema-grade visuals by simulating realistic light interactions with materials like leather, bronze, and metal, delivering dramatic improvements in realism and texture fidelity.

It’s not only geared for researchers but also for small teams and solo creators, with compatibility on consumer-grade GPUs, making high-end 3D asset creation broadly accessible.

Social Media Feature Updates and Algorithm Changes  

Video creation on Facebook just got easier 

Facebook is simplifying its video creation and sharing experience by making Reels the default format for all videos on the platform. Rolling out in the coming months, users will no longer choose between uploading a traditional video or a reel. Everything will now be shared as a reel, regardless of length or format.

Existing video content will remain on profiles and Pages, but any new videos created and posted moving forward will automatically be classified as Reels. As part of this transition, the “Video” tab on Facebook will be renamed to the “Reels” tab. 

Fediverse integration gets better on Threads 

Threads is significantly boosting its integration with the fediverse, an open network of interconnected social media servers. Users who have opted into fediverse sharing on Threads will gain access to two new features: a dedicated feed for content from federated users they follow, and the ability to search for fediverse users directly within Threads.

The new fediverse feed will display posts from platforms like Mastodon, Bookwyrm, and WriteFreely in reverse chronological order, allowing Threads users to see a broader range of perspectives without leaving the app. 

Additionally, users can now search for fediverse profiles by typing their full handle into the search bar, making it easier to discover and connect with individuals posting from outside Threads. 

WhatsApp introduces Subscriptions, Promoted Channels & Ads in Updates tab 

WhatsApp is no longer a simple messaging app. It has evolved into an effective channel for businesses to connect with their clients. Now giving it a whole new dimension, Meta is bringing monetization features within its Updates tab. 

The Updates tab currently houses both Channels and Status, aiming to help users discover new content and businesses without interrupting personal chats. These changes will be rolled out gradually over the coming months.

Notably, there are three new features. First, Channel subscriptions allow users to pay a monthly fee for exclusive content from their favorite channels.

Second, Promoted Channels will increase the visibility of channels in the directory to help users discover new interests.

Finally, Ads in Status enables businesses to promote products or services directly within Status updates, facilitating direct conversations with potential customers.

TikTok Symphony gets a fresh set of generative AI updates 

TikTok is enhancing its Symphony suite of generative AI tools, aiming to simplify and scale creative content production for brands, agencies, and creators. 

Unveiled at the Cannes Lions International Festival of Creativity, these new tools include Image to Video, Text to Video, and Showcase Products, designed to facilitate the creation and optimization of TikTok-first content.

Reddit rolls out major enhancements to comments on the platform 

Reddit is introducing a bunch of new features aimed at upgrading the user experience around comments. These enhancements include the ability to share comments as standalone posts, track real-time comment insights, and automatically save comment drafts.

Firstly, the “Share Comments as Posts” feature allows users to convert any comment into a fresh post to expand the discussion. Secondly, “Comment Insights” will now help Reddit users understand how their comments have performed in terms of metrics like upvotes, replies, views, shares, and awards. 

Finally, it’s now easier to engage with drafts of comments automatically saved on the platform, so that when you navigate away from the page and come back, you can pick up from where you left off. 

A Final Word…

With that, we bring the KIMP’s Picks June 20th Edition to a close. Keeping up with the overflowing updates on social media, scrolling through feeds to stay updated with all the latest buzz, can be tiresome. That’s why we curate this exclusive list of share-worthy posts at the end of every week so that you do not have to feel out of the loop. For more such updates, stay tuned, and we’ll be back next week with the next edition of the KIMP’s Picks.