KIMP’s Picks October 2nd: Weekly Highlights In Design, Marketing, AI, & More

Another week, another round of changes across AI, marketing, and social media. This week brings new AI models, smarter creative and advertising tools, fresh approaches to search visibility, and several updates from the major social platforms. Here’s a quick look at the developments that caught our attention.

Welcome to KIMP’s Picks’ October 2nd Edition.

In the Marketing Realm 

McAfee refreshes its brand as online security evolves 

McAfee has unveiled a new brand identity aimed at broadening how consumers perceive the company beyond its long-standing association with antivirus software. According to company research, 68% of consumers primarily recognize McAfee for antivirus protection, despite the company’s expansion into areas such as scam detection, identity monitoring, and dark web protection.

The rebrand reflects a broader shift in the cybersecurity landscape, where online threats are increasingly driven by sophisticated scams, identity theft, and AI-powered fraud. While McAfee continues to benefit from strong consumer trust built over decades, the company sees an opportunity to increase awareness of its wider portfolio of security services. 

McAfee’s rebranding effort underscores a larger trend across the technology sector, where businesses are working to align brand perception with the broader value they now deliver, rather than the products that originally made them famous.

Unique brand perspectives matter more for AI search visibility 

LinkedIn’s research into AI search argues that content structure alone is becoming less effective as a differentiator. While clear headings, lists, named entities, and data help AI systems extract information, according to LinkedIn, distinctive perspectives, original evidence, and expert insight are more important for earning citations.

Research cited by LinkedIn from Meltwater found that all 24 of the most-cited LinkedIn articles analyzed used bullet or numbered lists. The analysis also found that 93% had clear H2 or H3 headings, 75% named specific companies or tools, and 67% included hard numbers.

LinkedIn says AI systems tend to favor information that offers something specific to attribute, such as original research, proprietary data, first-hand experience, customer stories, or a verifiable industry insight. 

Google expands AI Mode monitoring to users worldwide 

Google’s AI Mode monitoring feature, which was initially available to Ultra and Pro subscribers, is now expanding to all users globally. 

The feature lets users tell AI Mode what information they want tracked. Google Search then continuously checks changing information across websites, forums, social posts, real-time data sources, and its Shopping Graph, which contains more than 60 billion products.

Users can monitor topics such as new restaurants and pop-ups, local activities, product restocks, and price drops. Google says AI Mode can also suggest monitoring tasks while users search, helping them keep track of information that changes over time.

AEO and SEO converge as AI changes how people search 

Search visibility is expanding beyond traditional Google results as users increasingly turn to AI-generated answers from platforms such as ChatGPT, Perplexity, and Google’s AI features. The key distinction is that SEO focuses on organic search rankings, while AEO aims to get content cited or mentioned in AI-generated answers.

The two approaches share important foundations. Research cited by Semrush suggests AI systems often favor authoritative domains with strong backlinks and organic visibility, although they may pull different pages from those domains than the ones ranking highest on Google.

The article recommends creating original, research-backed content, building backlinks and brand mentions, using consistent entities and terminology, and structuring pages with clear headings and direct answers. Keeping information genuinely updated is also important, particularly for topics that change quickly.

For AEO, content should be easy for AI systems to extract and understand. Short, standalone answers and question-based sections can help systems identify relevant passages for citations. The article also recommends tracking traditional rankings alongside AI visibility, brand mentions, citations, and the prompts where a brand appears.

Time for small businesses to rethink social media strategies 

Entrepreneur magazine recently shared a post highlighting changes in consumer behavior and why small businesses need to stop treating social media like a digital billboard and start operating more like local media brands. 

Traditional tactics built around polished images, promotional graphics, and direct sales messages are becoming less effective as consumers seem to now favor authentic, people-first content.

The article highlights a growing preference for short-form video, behind-the-scenes content, and real customer interactions. Platforms such as YouTube, Facebook, Instagram, TikTok, and Reddit continue to attract large audiences, but discovery-driven formats like Reels, Shorts, and TikTok videos are playing a bigger role in helping businesses reach nearby customers.

The focus is also moving away from vanity metrics like likes and follower counts toward outcomes that directly impact business performance, including inquiries, appointments, and sales. 

In the AI Realm 

Google unveils Gemini 4 Argon with focus on deep reasoning and cybersecurity 

Google has announced Gemini 4 Argon, a new frontier AI model designed for deep reasoning across complex workflows. Currently rolling out through the Fairwind Program, Argon reportedly excels in real-world software engineering, finance, legal tasks, and cybersecurity defense.

The model introduces an industry-leading output token limit of 1 million, enabling it to handle long, multi-step tasks in one go. Google reports major internal productivity boosts, including fixing data center memory efficiency and accelerating code migrations. Furthermore, Argon features specialized cybersecurity capabilities to help defenders autonomously find, validate, and patch critical software vulnerabilities. 

Before a broader release to developers and enterprises, Google is actively testing safety guardrails, including defenses against prompt injection, monitoring for misalignment, and hardening sandboxed environments.

ChatGPT Space simplifies team collaboration and AI-powered workflows 

OpenAI has launched ChatGPT Space, a new collaborative workspace designed to help teams work alongside AI in a shared environment. The feature introduces Pages, which are interactive documents that can include charts, images, checklists, dashboards, and other dynamic content.

ChatGPT can generate Pages based on a user’s conversation or task requirements, helping teams organize information, track projects, and manage ongoing work. Users can edit Pages collaboratively and tag ChatGPT, Codex, or teammates to assist with action items, updates, and workflow management.

The platform also supports connected data sources, so Pages can stay updated with relevant information as work progresses. The goal, reportedly, is to make AI a more active participant in team collaboration rather than a standalone chat interface.

Gemini rolls out “Skills” to replace Gems for task automation 

Google is introducing “Skills” in Gemini to help users simplify task automation and avoid the need for lengthy prompts for repetitive tasks. By saving custom instructions once, you can reuse them instantly in the prompt bar by typing a forward slash.

This new feature lets you prep for presentations, match your personal writing style, or gather multiple perspectives for brainstorming. You can even stack multiple skills together and attach reference files like PDFs and images for more tailored results. 

The launch also marks the beginning of Google’s transition away from Gems. Over the coming months, Skills will gain capabilities currently available in Gems, including sharing and integrations with Google Drive and Gemini Notebook.

OpenAI now announces shareable profiles in ChatGPT 

OpenAI is rolling out shareable ChatGPT profiles that let users create a public-facing profile with their name, photo, username, bio, activity, and selected Sites. Personal profiles remain private by default, while users can choose to make them public and share a profile link.

Public profiles do not expose conversation titles or chat content. Users can also customize their profile’s layout, theme, and visible sections. Profiles can display Chat message counts, Work and Codex token usage, activity streaks, and top plugins, with activity shown separately for each workspace.

The update also adds a Site showcase, allowing users to feature up to 12 Sites they own. The visibility of each Site remains controlled by its existing permissions, so adding one to a profile does not change who can access it.

Anthropic releases Claude Sonnet 5.5 with faster performance and lower costs 

Anthropic has released Claude Sonnet 5.5, the second model in its Claude 5.5 family, positioning it as a faster, lower-cost option for everyday professional work. The model is designed for tasks including coding, bug fixes, document creation, presentations, and spreadsheets.

Anthropic says Sonnet 5.5 generates responses more than 30% faster than Sonnet 5 and can cost up to 30% less per task because it typically uses fewer tokens. 

The model also shows major gains on several benchmarks. According to Anthropic, the model scored 70.6% on Terminal-Bench 4.0, compared with 10.3% for Sonnet 5, and reached 1844 on GDPval-AA, nearly matching Opus 5.5 at 1846.

Sonnet 5.5 is built to improve long-horizon tasks, visual understanding, computer use, and professional knowledge work. It also includes stronger cybersecurity safeguards and protections against capability extraction.

OpenAI announces new always-on AI agents called Dots 

OpenAI has introduced Dots, AI agents designed to work continuously on tasks using their own cloud computers, browsers, and connected apps. Powered by GPT-6 Astra, Dots can learn user preferences from feedback and work toward goals with less step-by-step direction.

Dots can connect with more than 4,000 apps through OpenAI’s plugin ecosystem and can be accessed through ChatGPT, Slack, and Teams. They can handle tasks such as software development, updating launch materials, analyzing new research, preparing sales proposals, and producing content.

The agents can also work in the background through a feature called proactive research. These background tasks use read-only access to connected apps unless users provide additional permissions. Users can set custom rules, monitor activity, and approve actions that could affect accounts or share information.

Runway introduces a new AI-powered ad platform 

Runway has unveiled Runway Ads, an autonomous tool designed to manage and optimize paid advertising creative from start to finish. By linking an ad account and brand guidelines, the platform automatically generates video and image assets, publishes approved variations directly to Meta, Google, and TikTok, and analyzes performance data to produce the next wave of high-performing ads.

In performance marketing, continuous testing and high creative volume are crucial, yet most teams struggle to produce enough content variations to reach peak campaign efficiency. Runway Ads eliminates this creative bottleneck for organizations of all sizes. Enterprise brands can rapidly run global, localized campaigns, while smaller teams and creators can scale up testing with thousands of weekly ad variations without taking time away from their core operations.

Social Media Feature Updates and Algorithm Changes

Forum from Meta is a dedicated app for Facebook Groups 

Meta had quietly soft-launched a Reddit-like app for communities called Forum. Now the platform has announced an official rollout. 

The Forum app is designed to make Facebook Groups easier to browse and participate in. The app syncs with users’ existing Facebook groups and also helps them discover public communities based on their interests. Forum has been available for testing since May and is currently available in the US on iOS and Android.

The company is also testing new tools focused on community participation. A new recognition role highlights members who regularly provide useful answers, firsthand experiences, practical recommendations, and conversations that help their groups. Admins can assign the role, while Meta can also nominate members.

Forum’s Ask feature uses AI to surface relevant posts and comments from groups, helping users find advice based on real experiences. New features can also compare recommendations and suggest questions based on recent searches.

Muse for Small Business is built for small business operations 

Meta is expanding its Muse AI agent with new skills and connectors designed to help small businesses manage everyday work. Muse can now connect with tools including Shopify, QuickBooks, Canva, Slack, Stripe, Asana, Notion, Klaviyo, Zoom, and business accounts on Facebook and Instagram.

Once connected, Muse can work with business information such as products, brand details, customer questions, sales data, and marketing activity. Businesses can use it to analyze performance, review expenses, improve advertising and content, organize schedules, and draft responses to emails.

Additionally, Muse can also proactively flag tasks that need attention and suggest ways to address them. 

According to Meta, users remain in control. Muse cannot publish content, send communications, or spend money without approval.

View on Threads

Meta unveils Edits Assistant 

Meta is officially launching its AI-powered Edits assistant to all users in the US. Unlike generic AI tools, this assistant connects directly to your Instagram account to provide personalized insights based on your specific performance data.

The tool analyzes key metrics such as views, retention, likes, and comments, alongside platform trends to pinpoint exactly what resonates with your audience. Creators can use it to brainstorm video concepts, discover niche trending audio, refine captions, and uncover performance patterns over time. Developed with direct feedback from top creators, the assistant focuses on heavy data analysis, leaving full creative control in the hands of the user.

Edits assistant is free to access within the Edits app with daily prompt limits, and expanded access will soon be available through a Meta One subscription.

Meta expands school partnership program with new student features 

Meta is expanding its School Partnership Program to help schools keep students and parents informed on Instagram. The updated program introduces verified school hubs, letting administrators share official updates through broadcast Channels, School Stories, and Notes.

To create a safer environment, Meta partnered with third-party verifier UNiDAYS to confirm high school student identities before granting access. Verified students can join a dedicated school directory, showcase a school banner on their profile, and follow specific Clubs and Teams. School administrators maintain full moderation control to remove posts or accounts, while safety reports from partner schools receive prioritized reviews. Parents can also track their teen’s school community involvement through Meta’s Family Center.

These updates aim to streamline school communications within a secure digital space outside classroom hours.

Wrapping Up…

What happens when AI search, smarter agents, automated advertising, and social media updates all move forward at once? You get another busy week in the digital world. And keeping up feels overwhelming. That’s where our weekly social media recap comes in. Instead of spending all weekend doomscrolling just to catch up on all that you missed, get your bite-sized updates right here. Or better yet, subscribe to The KIMP Scoop newsletter and get your quick recap delivered to your inbox.