KIMP’s Picks June 5th: Weekly Highlights In Design, Marketing, AI, & More

Every day you open social media, there are a bunch of new updates to catch up. These could be AI news that’s sending ripples through the industries or AI-focused announcements from the top tech firms. These could also be about how your favorite social media platforms are changing.

This is news that might change the way you work, change the way you find and consume content or even impact your productivity in the long run. Want a quick way to catch up on all such highlights that matter? Welcome to KIMP’s Picks, 5th June edition.

For all those marketers and entrepreneurs looking to stay ahead in their game, let’s start with some marketing inspiration, useful industry insights, and a quick recap of the recent AI news and more. 

In the Marketing Realm 

Google reminds us that human creativity still leads 

While one part of the world is debating the need for collaborating with human designers and considering using AI tools to generate videos, Google recently released a video that shows how their team is using AI. And this behind-the-scenes video flips the script. 

Behind the scenes of their “Timmy TPU” project, Google demonstrated that the future of content creation isn’t a race to replace human artists, but all about collaborating with them efficiently. 

Instead of generating a film from purely textual prompts, creators built a physical, lo-fi world out of cardboard, pipe cleaners, and traditional puppets. Feeding these real-world puppetry performances into models like Gemini Omni, Google also ensured that the result preserved the human touch. It retained the charming, shaky, and imperfect nuances of the puppeteers’ hands while instantly elevating the visual style into a cinematic masterpiece.

For brands and marketers, this is a massive wake-up call. The real value of generative AI lies in its ability to act as a creative partner who helps save hours without sacrificing the authentic, tactile human heart of the story. In an era of AI fatigue, the brands that stand out will be those that follow Google’s lead, using cutting-edge technology not to eliminate the artist, but to amplify the craft.

Google unveils major updates across Ads and other tools in Google Marketing Live 2026

At the Google Marketing Live event, Google highlighted how AI is reshaping advertising across Search, creative production, and measurement. Leaders from Google Ads shared how Gemini models are now central to understanding user intent and powering new ad systems built for more complex search behaviour.

As you know, Search is changing fast. Queries are getting longer, more conversational, and increasingly multimodal, with more users adding follow-up questions and using visual inputs. Google says that this makes keyword-only targeting less effective. 

Instead, systems like AI Max and Performance Max now rely on broader intent signals, keyword expansion, and AI-generated ad copy that adapts to each query. Advertisers can still maintain control through AI Brief, which lets them set brand rules and messaging direction in natural language.

These changes highlight a shift toward AI-led targeting, faster creative testing, and more predictive measurement to stay competitive in evolving search environments.

Trust is a growth lever that marketing teams seem to be ignoring 

A recent post from ADWEEK talks about a crucial factor that marketing teams seem to be ignoring – trust. Marketing has become highly optimized, but many brands are still struggling to grow. The core issue is not execution or targeting. There is a growing trust gap between customers and brands.

Over the past decade, marketing has leaned heavily into data, funnels, and performance metrics. Brands now know exactly who to target and how to track conversions. But despite all this precision, customers are easier to ignore than ever. The problem is not media efficiency. It is a lack of emotional connection.

Much of modern marketing focuses on measurable actions like clicks and conversions. But purchase decisions are driven first by emotion, then justified by logic. Trust plays a far bigger role in buying behaviour than most strategies account for. 

Personalization is not enough. Tailored messaging can also sometimes feel automated. Customers respond more strongly to human moments of care and memory than to scaled messaging systems. Real trust is built outside dashboards. As automation and AI make marketing more uniform, trust becomes the only real differentiator.

Marketing leverage and why teams should focus on it 

Marketing is ultimately about movement. Taking people from awareness to curiosity, then to confidence, action, and ideally long-term advocacy. The real force behind that movement is leverage, not budget or output.

The post here from MarTech talks about marketing leverage. Many teams mistake activity for leverage. They invest in complexity, assuming clarity exists on the customer side when it often does not. Others rely too heavily on data without a real understanding, or confuse past experience with current market needs.

Real leverage comes from better customer understanding. That means direct conversations, observing actual behaviour, and combining signals from analytics, reviews, and frontline teams. Growth depends less on louder marketing and more on a clearer understanding that genuinely moves people to act.

In the Content Marketing Realm 

Google Search Console adds a generative AI report for AI Overviews and AI Mode visibility 

Google has started rolling out a new Generative AI report inside Search Console, giving site owners visibility into how their content appears in AI Overviews, AI Mode, and AI-driven features in Discover. This is the first direct way to see how often pages are being surfaced in Google’s generative search experiences.

The report is currently being introduced to a limited group of users, mainly in the UK, with a wider rollout expected later. It sits under the Performance section and shows impressions generated from AI features. While it does not include clicks, rankings, or CTR, it still gives clear insight into which pages Google’s systems consider relevant for AI responses.

This means better clarity for brands on how SEO performance extends into AI-driven search experiences and where content needs to evolve to stay visible.

In the Design Realm 

Canva comes to Perplexity Computer 

Perplexity has partnered with Canva to bring design creation directly into its Perplexity Computer platform. This closes the gap between AI-generated ideas and finished visual assets. The integration allows users to move from research and strategy outputs straight into editable Canva designs without switching tools.

The update is designed for small and growing businesses, where speed and limited resources often make it difficult to turn ideas into polished presentations, campaigns, or client-ready materials. 

With this integration, Perplexity can generate structured briefs using live data, meeting notes, and web context, which Canva then transforms into formats like pitch decks, social media content, infographics, and brand kits.

Brand identity built around a human-centered visual system 

The Brand Identity recently shared a post discussing the new brand identity for the Creators Coalition on Artificial Intelligence (CCAI). The identity was designed to feel grounded, flexible, and rooted in community rather than corporate tech aesthetics. The project aims to reflect a human-centred stance on AI while remaining adaptable across formal and expressive use cases.

At the core of the identity is a logomark made of four intersecting circles representing the letters C, C, A, and I. A spark-like symbol emerges where the circles meet, chosen deliberately despite its heavy use in AI branding. The studio reframed it as a shared visual cue rather than a corporate cliché, using it to represent connection and collective action. 

A key part of the identity is a custom web tool that lets users create textured, hand-like visuals using the palette. This turns the brand system into something participatory, where the community can actively shape its expression rather than just consume it.

This brand identity project demonstrates the growing shift toward identity systems that prioritize participation, transparency, and human expression over static corporate branding.

In the AI Realm 

Google announces Gemma 4 12B 

Google has introduced Gemma 4 12B, a new mid-sized AI model designed to deliver multimodal, agentic capabilities directly on consumer laptops. It sits between the lightweight edge-focused E4B model and the larger 26B Mixture of Experts version, offering a balance of performance and efficiency with a reduced memory footprint.

A key milestone is native audio input support, making this the first Gemma model in its range to process voice directly. Combined with vision capabilities, the model is built to handle multimodal tasks without relying on separate encoders. This helps reduce latency and improve efficiency.

Despite its smaller size, the model delivers performance close to the larger 26B version on benchmarks, while being light enough to run locally on devices with around 16GB of memory.  

Android June Drop expands AI-powered upgrades to more phones 

Google has rolled out its latest Android June feature update, introducing a mix of safety, AI-powered tools, and improved sharing across devices. 

A key safety upgrade is fake call detection in the Phone by Google app, which can now flag suspicious calls where scammers attempt to impersonate trusted contacts.

On the discovery side, Circle to Search is expanding its capabilities, allowing users to identify full outfits directly from images without switching apps. This makes visual search more detailed and practical for shopping and style inspiration.

Google Photos is also adding a digital wardrobe feature that organizes clothing from user photos into a virtual closet. 

Google’s Dreambeans experiment turns data into daily story feeds 

Google Labs has introduced Dreambeans, a new AI experiment designed to generate personalized daily story collections using its latest “Personal Intelligence” system and Nano Banana 2 model. The feature aims to reduce information overload by replacing endless scrolling with curated, finite story experiences.

Dreambeans connects with Google apps like Gmail, Calendar, Photos, YouTube, and Search history to build context-aware stories tailored to each user. These stories are designed to feel relevant and timely based on upcoming calendar events.

Each story includes custom AI-generated illustrations and can be expanded for deeper exploration, pulling in additional web information like local recommendations or relevant learning resources. 

The experiment is currently rolling out to Google AI Ultra subscribers in the US and users in other countries can join the waitlist. 

Seven new AI models added to the MAI family 

At Microsoft Build, Microsoft AI announced a new family of seven in-house models under its MAI (Microsoft AI) umbrella, alongside a broader vision for a “superintelligence lab” focused on advancing frontier AI systems for real-world use.

The MAI model suite spans image, voice, transcription, coding, and reasoning tasks, designed to work together as a multimodal ecosystem. Key releases include MAI-Thinking-1 for advanced reasoning, MAI-Code-1-Flash for efficient agentic coding, MAI-Image-2.5 for text-to-image generation and editing, MAI-Transcribe-1.5 for high-accuracy speech recognition, and MAI-Voice-2 for natural speech synthesis across multiple languages.

Microsoft also highlighted early work with Mayo Clinic to develop a specialized healthcare AI model designed to support clinical reasoning, diagnostics, and treatment planning within secure environments.

Plugins, Sites and more new updates coming to Codex 

OpenAI is expanding Codex beyond software development into broader knowledge work. OpenAI is introducing new features designed for analysts, marketers, designers, investors, and other non-technical users. The platform now serves over 5 million weekly users, with non-developers making up around 20 percent and growing at more than three times the rate of developers.

The update introduces role-specific plugins that connect Codex to tools and workflows across different industries. These include analytics integrations for business intelligence tools, creative production support for design platforms, sales workflows linked to CRM systems, and finance-focused plugins for investing and banking tasks. Each plugin bundles apps, skills, and instructions tailored to specific job functions.

Codex is also introducing “Sites,” a new feature that allows users to turn prompts into interactive, shareable web apps and dashboards. These can be used for project tracking, customer reviews, financial planning, and internal collaboration, all accessible via a simple URL. 

Microsoft introduces Scout, an always-on AI agent 

Microsoft has introduced a new category of AI agents called Autopilots, designed to operate continuously in the background and carry out tasks across workplace systems without requiring repeated prompts. These agents are built to maintain context, follow defined permissions, and keep work progressing across apps and platforms.

The first Autopilot agent is Microsoft Scout, integrated across Microsoft 365 tools, including Teams, Outlook, OneDrive, and SharePoint. It connects cloud, desktop, and web environments, allowing it to coordinate tasks, manage schedules, surface important meetings, and prepare relevant materials automatically.

Unlike traditional assistants that respond to individual requests, Scout is designed to maintain context over time. It can proactively identify meetings, deadlines, and deliverables, block calendar time, and flag potential risks such as stalled decisions or missing inputs. 

Ideogram’s new text-to-image model brings advanced design control 

Ideogram has released Ideogram 4.0, its first open-weight foundation model for text-to-image generation. The model is built as a 9.3 B-parameter Diffusion Transformer trained from scratch. It is designed to push open-source design AI forward with stronger layout control, typography handling, and structured prompting.

Ideogram 4.0 is optimized for high-quality image generation and editing. Instead of free-form text prompts, the model is trained exclusively on JSON-based inputs that define composition, styling, color palettes, and even bounding boxes for precise layout control. This allows more predictable and controllable image generation compared to traditional prompt-based systems.

It adds advanced text rendering, editable text, movable layers and other useful features for better control over the design output. 

Social Media Feature Updates and Algorithm Changes 

Meta Business Agent helps automate customer conversations 

Meta has introduced Meta Business Agent, an AI system designed to help businesses handle customer interactions at scale across WhatsApp, Messenger, and Instagram. The tool is built to act like an always-available digital team member, responding to customers, managing conversations, and supporting sales and service workflows.

The agent can be set up quickly and integrated into existing business systems. It is already being used by over one million businesses and operates across more than a billion daily business messaging threads, helping deliver faster and more personalized responses in real time.

Meta Business Agent can answer product questions, recommend items from catalogs, book appointments, and also provides businesses with a morning briefing feature that summarizes missed messages and highlights key insights from customer interactions.

Threads introduces music in posts 

Meta’s Threads is rolling out a new music feature that allows users to add songs directly to posts and replies. The update is designed to make expression on the platform more engaging by blending text, visuals, and audio in a single post.

Users can tap the music icon in the composer, search for a track, or pick from trending songs. Once selected, they can choose a 30-second clip to highlight within their post.

The feature also includes flexible display options, letting users switch between lyrics and album artwork or keep a simplified view showing just the song and artist. Posts can also include photos or videos alongside the music, giving users more creative control over how they share content.

View on Threads

Meta expands teen content controls across its platforms 

After rolling out paid “Plus” subscriptions last week, Meta is rolling out expanded safety updates for Teen Accounts across Instagram, Facebook, and Messenger, strengthening default content protections for users under 18. The update focuses on limiting exposure to mature content and improving how teens experience recommendations across feeds, Reels, and messaging.

The 13+ content setting, which was first introduced in select countries, is now expanding globally. It is designed to automatically filter out inappropriate content and restrict interactions with accounts that frequently share age-restricted material. A stricter “Limited Content” mode is also being made available for families who want tighter controls.

Meta is also testing a new system that reduces repeated exposure to similar types of content. This is aimed at improving content variety so teens are not repeatedly shown the same themes, even when the content itself is considered safe or educational. 

X will now let you share video responses on the iOS app 

X has launched a new “React with Video” feature for iOS users. This allows people to reply to posts using short video responses instead of text. The update is designed to make commentary more visual and interactive, building on the platform’s push toward creator-led engagement.

Users can respond through the repost option and record video reactions using formats like green screen, split screen, or picture-in-picture. This makes it easier to add context, commentary, or storytelling directly onto existing posts.

Another platform that is joining the video responses trend is Reddit 

Reddit is rolling out a new feature that allows users to respond to comments with videos directly inside threads. The update is designed to make discussions more dynamic by adding video alongside traditional text, images, and GIF replies.

The feature will launch globally on June 11 in all public, safe-for-work communities, following an early rollout to moderators starting June 3. This phased approach gives community moderators time to adjust settings before full deployment.

Video replies can be used for tutorials, AMAs, reactions, screen recordings, or more creative responses. Each comment supports one video along with optional text, and videos will play inline across desktop and mobile without autoplay. 

Wrapping Up…

On one side, we’re seeing AI models get closer to real utility on everyday devices. We’re seeing more new, smaller, faster, and more capable systems that don’t need massive infrastructure to be useful. Intelligence is moving off the cloud and into the tools people actually hold and use. But staying up to date with all these highlights can be overwhelming and that’s where our weekly recap post comes in. For more updates, stay tuned and we’ll be back with the next edition of KIMP’s Picks soon.