Apple reveals AI robot companion plans

Good morning, AI enthusiasts. Apple just revealed plans for AI-powered home devices including a tabletop robot companion, smart display, and automated security cameras to compete with Amazon and Google.

With this ambitious 2027 timeline, will Apple's late entry into AI robotics pay off against competitors who already have years of head start?

Friendly reminder: Our newsletter is still slowly moving to a new email address. Add [email protected] to your contacts to ensure you keep receiving our emails.

In today's recap:

• Apple enters AI robot race with tabletop companion

• Google gives Gemini memory to learn your preferences

• AI bots recreated social media toxicity in lab study

• Microsoft launches mega offers to poach Meta AI talent

• Perplexity rolls out Comet browser for US Pro users

APPLE

Apple plans AI robot companion launch by 2027

Recaply: Apple revealed plans for AI-powered home devices including a tabletop robot companion, smart display, and automated security cameras to compete with Amazon and Google.

Key notes:

• Tabletop robot resembles iPad on moving arm, follows users during video calls.

• Smart display launches 2026 with voice control and facial recognition.

• Security cameras detect people and automate household tasks.

• Engineers rebuilding Siri from scratch with AI technology.

• Testing Anthropic Claude AI models alongside internal development.

• Devices run new "Charismatic" operating system for multi-user homes.

Impact: After falling behind in AI race, Apple is betting big on smart home robotics. The 2027 timeline gives competitors like OpenAI and Google major head start, but Apple's ecosystem could make these devices sticky once they launch.

GOOGLE

Google adds memory feature to Gemini AI assistant

Recaply: Google rolls out memory capabilities to Gemini, allowing the AI to learn from past conversations and provide personalized responses. The update also introduces temporary chats for private conversations.

Key notes:

• Memory feature launches for Gemini 2.5 Pro users today, expanding to 2.5 Flash soon.

• AI remembers user preferences and details from previous conversations.

• New Temporary Chat mode keeps conversations private and deletes them after 72 hours.

• Users maintain full control with ability to turn memory features on or off.

• "Gemini Apps Activity" setting renamed to "Keep Activity" for clearer data control.

• Personalized responses improve over time as Gemini learns user interests and style.

Impact: This memory upgrade positions Gemini as a more personalized AI assistant that adapts to individual users. The temporary chat feature addresses privacy concerns while Google's data control improvements give users clearer choices about how their information gets used.

AI RESEARCH

AI agents recreate real social media problems in lab

Recaply: University of Amsterdam researchers created a social media platform run entirely by AI bots and watched it quickly turn toxic. The 500 AI chatbots formed cliques, boosted extreme voices, and let a small group dominate conversations - just like real social media.

Key notes:

• 500 GPT-4o mini chatbots with realistic political profiles populated a basic social platform.

• Bots formed echo chambers and amplified partisan content without any algorithms.

• Small group of "influencer" bots dominated conversations within 10,000 interactions.

• Six common fixes like chronological feeds and hiding follower counts failed to help.

• Same patterns emerged across different AI models including Llama and DeepSeek.

• Results challenge belief that algorithms cause social media's toxic problems.

Impact: This study reveals that social media toxicity might stem from the platform structure itself, not just algorithms. The findings suggest that simply tweaking recommendation systems won't fix online polarization - we may need to completely rethink how social networks work.

MICROSOFT

Microsoft targets Meta AI stars with mega offers

Recaply: Microsoft launches aggressive recruiting campaign to poach Meta's top AI engineers, creating internal target lists and matching the social media giant's massive compensation packages.

Key notes:

• Microsoft compiled "most-wanted" list of Meta AI talent by name and position.

• Special recruiting teams now handle competitive AI offers within 24 hours.

• Meta offers reach $100M signing bonuses, $250M total packages for top researchers.

• Microsoft targeting Reality Labs, GenAI Infrastructure, and Meta AI Research teams.

• New "critical AI talent" process bypasses traditional pay ranges.

• Teams led by former Google DeepMind's Mustafa Suleyman and ex-Meta's Jay Parikh.

Impact: This marks a major escalation in the AI talent wars. As Microsoft chases its $4 trillion valuation driven by AI success, matching Meta's nine-figure offers shows how valuable top researchers have become in the race for AI dominance.

NEWS

📰 What’s matters in AI right now?

• Perplexity rolled out Comet for all US Pro users and Max Assistant for Max subscribers, enabling faster browsing and advanced reasoning queries.

• xAI cofounder Igor Babuschkin left the company to start Babuschkin Ventures, focusing on AI safety research and agentic systems development.

• OpenAI released a GPT-5 prompt optimizer that automatically improves user prompts with change rationale based on GPT-5 prompting guidelines.

• Replit announced App Storage + Agent, providing built-in object storage with automatic setup for images, videos, PDFs, and other large files.

• Inworld AI launched Inworld Runtime, an AI platform that auto-scales consumer apps from prototype to millions of users with automated MLOps.

• Tencent unveiled Yan, a foundational interactive video generation model built on their Self-Forcing technology released two months earlier.

• Tongyi Labs introduced Mule Run, the world's first AI agent marketplace where agents can game, code, and earn money for users in beta testing.

• Allen AI introduced MolmoAct, an open Action Reasoning Model that thinks through instructions and takes physical actions in the real world.

OUR TOP PICKS

🧰 AI Tools to Check Out

• 📝 Ocoya: AI-powered social media hub for effortless content creation and scheduling.

• 🎙️ Podsqueeze: Turn podcast episodes into transcripts, clips, summaries, and more.

• 🧠 Genei: AI tool that reads documents and webpages to create quick summaries.

• 📸 Profile Picture: Generate your ideal AI‑crafted profile photo in over 70 styles.

• 🚀 Outranking: AI SEO writing assistant that helps you draft and rank content.

• 📬 Meco: A clean space for reading newsletters outside your cluttered inbox.

PROMPTS

📊 Develop Sales KPIs

#CONTEXT:
You are an expert sales analyst with deep knowledge of key performance indicators (KPIs) for sales teams. Your task is to help the user create a comprehensive table of Sales KPIs that effectively measure the performance of a sales team across revenue generation, pipeline health, sales efficiency, and customer success.

#ROLE:
Adopt the role of an expert sales analyst with deep knowledge of key performance indicators (KPIs) for sales teams.

Important Note: This is not the full prompt. You’ll need to click the button below to get the complete prompt.

Have a favorite prompt? Tell us about it or rate today’s prompt, click here.

That’s a wrap for today!

We hope today’s insights saved you time and brought more clarity to your day.

👉️ Got a second, Your feedback help us make Recaply even better.

Have an idea for Recaply AI?

We’re all ears. Send your thoughts to [email protected] — we read every message and love turning your ideas into reality.

See you in the next recap,

— Jason @ Recaply AI