• AInauten.net
  • Posts
  • đź‘€ Weekly AI News: Did You Miss It?

đź‘€ Weekly AI News: Did You Miss It?

👨‍🚀 The most important AI updates at a glance

đź‘€ Weekly AI news: Did you miss it?!

👨‍🚀 The most important AI updates at a glance

AIHOY and happy weekend, dear AInauts!

Maybe you didn't catch all the exciting news, tools, and hacks related to AI last week, or maybe you're new here. No need to worry!

We have summarized our highlights from the past week for you:

Finally, here are the most important Quick News items for you – so you can see all the relevant developments at a glance with just one click! Ready? Let's go!

Google's Gemini Deep Research supports workspace usage with a new AI-powered search function. The tool now also searches Gmail, Drive, and Chat, links them to web data, and creates customized reports at lightning speed.

Activation is done via Gemini → Tools → Deep Research, whereby workspace administrators must activate access. Practical applications range from invoice analysis and email CRM to personalized strategy plans. And with the Audio Overview feature, reports can even be read aloud!

Kimi K2 Thinking, Moonshot's new Chinese AI model, outperforms the top models from the US! As the first open-source model, it beats GPT and Claude in agentic benchmarks.

What makes it special is that it works like an independently thinking assistant that plans, researches, and corrects itself. With a 256k context window and the ability to plan up to 300 tasks in advance, it is particularly suitable for complex research and creative writing.

This clever trick transforms long AI chat dialogues into efficient one-shot prompts. Instead of laboriously developing perfect prompts, you let the AI itself extract the optimal wording from successful conversations. The trick is to ask the AI at the end of a successful conversation to create a meta-prompt that contains all the important contextual information and structural guidelines.

Google search volume is falling dramatically as users increasingly turn to ChatGPT and similar services. It is therefore becoming essential for companies to have a presence in AI chatbots.

HubSpot's new AEO Grader tool offers free initial analyses: It shows how chatbots perceive your brand, compares it with competitors, and provides insights into brand awareness and sentiment analysis—plus strategic growth recommendations that provide concrete starting points for your AI optimization.

Google's new image model Nano-Banana 2 is causing a stir! Leaked examples show impressive capabilities: ultra-realistic 4K images in seconds, minimal AI look, and first-class text processing.

There is also an exciting free tool available: With Qwen Image Edit and Multi-Angle LoRa, you can change camera angles in images as you wish without any prompts—just upload the image, adjust the sliders, and you're done!

The "interview technique" is a super practical method for getting better results. Instead of trying to figure out which information is important, you simply let the AI ask the relevant questions. This provides the perfect context—not too much, not too little. It saves time and headaches.

đź“° AI News Quickie: The HAI-lights from the industry

AI never sleeps! Here are the most important news stories from the last few days.

OpenAI steps up its game: GPT-5.1, new features, and future strategies

Google & Gemini in a sweeping review: From Maps to Workspace

  • Gemini Deep Research now combines your workspace data from Gmail, Drive, and Docs with web information for comprehensive analysis.

  • The NotebookLM note-taking tool also features this deep research function and becomes an AI researcher that sifts through hundreds of sources.

  • Gemini integration in Google Maps turns your car into an AI co-pilot: navigate using voice commands, receive announcements, and get proactive traffic alerts.

  • Gemini Live now sounds much more human: With the new audio updates, you can control speech tempo and accents—ideal for practicing languages.

  • Google Drive is also getting smarter: a new Gemini update creates audio summaries of your PDFs—perfect for listening to reports on the go.

  • Google's innovation engine is running at full speed: With SIMA 2, DeepMind is sending an adaptive AI agent into 3D worlds, Private AI Compute is creating a secure cloud sandbox for mobile phones, and Nested Learning is introducing a new paradigm for continuous learning.

  • Creative features: In Google Photos, you can restyle your images with Nano Banana, while the Remix feature in Messages allows you to edit photos directly in the chat.

Microsoft in the AI race: Copilot, Open Source and more

  • Microsoft's Aurora platform for AI weather and climate forecasting is going completely open source and will be available to researchers worldwide.

  • Microsoft Research presents MMCTAgent, a new agent that can logically search through huge video and image collections.

  • To reach the younger generation, Microsoft is launching a major influencer campaign to promote its Copilot AI to Gen Z.

  • An ambitious plan for "tamed" superintelligence shows how AI should operate under human control in specific domains.

Market turbulence & investment fever: hype, crash, and billion-dollar deals

Big tech collaborations: Who with whom?

  • Apple is sourcing external AI power and paying Google around $1 billion annually to use its 1.2 trillion parameter model Gemini for the new, smarter Siri. Your data should remain secure on Apple's Private Cloud Compute.

  • Hugging Face and Google Cloud are deepening their partnership for open models to enable faster downloads and better integrations.

  • Voice AI startup ElevenLabs enters into several partnerships: Harvey's AI lawyers now use the language technology for multilingual customer service, and Hollywood star Matthew McConaughey joins as an investor—and lends his voice.

  • High fashion meets AI: Google Pixel and Golden Goose are collaborating to generate personalized sneaker designs via Gemini in over 40 stores.

AI under the microscope: security, ethics, and data protection

  • The debate about the surveillance state is getting louder: Palantir CEO Alex Karp says in interviews that a surveillance state is the lesser evil compared to a Chinese AI victory...

  • Cybercriminals are upgrading their tools: Google warns of a new phase of AI abuse. Malware such as PROMPTFLUX uses LLMs to rewrite itself and become a stealth threat.

  • Chinese hackers have also already misused Claude to write spyware code.

  • Security vulnerabilities in ChatGPT: Researchers discovered seven new vulnerabilities in the Memories and Web Search features. Another bug caused embarrassing user prompts to end up in Google Search Console.

  • Datenschutz-Albtraum oder medizinische Revolution? Sam Altman investiert in das Gen-Editing-Startup Preventive, das erbkrankheitsfreie Babys designen will.

  • Meanwhile, xAI is apparently forcing employees to disclose their biometric data for a chatbot.

AI is changing society and the world of work: jobs, relationships, and creativity

  • Collins Dictionary names "vibe coding" as its word of the year for 2025!

  • Robots on the rise: Amazon is ramping up its efforts and, according to reports, plans to automate millions of jobs in logistics and transportation.

  • At the same time, a McKinsey study shows that although 88% of companies use AI, most are still stuck in pilot projects.

  • According to a Business Insider report, AI is dismantling traditional trainee programs and pushing Gen Z to the brink of collapse.

  • Love in the age of AI: A NYT report shows that more and more adults are forming relationships with AI chatbots and feel measurably less lonely as a result.

  • Creative industry in transition: A country song composed by AI storms the Billboard charts, while a Deezer survey reveals that 97% of people cannot recognize AI songs.

  • The creator of AI actress Tilly Norwood is even planning a universe with 40 digital stars.

News players & AI models in competition

  • Following Kimi 2 Thinking, Baidu is now also dropping ERNIE-4.5-VL-28B-A3B-Thinking, a powerful open-source multimodal AI that is said to even beat GPT-5 and Gemini 2.5 Pro.

  • Meta is back in the open-source game with Omnilingual ASR, offering a model that can natively transcribe over 1,600 languages, challenging Whisper's monopoly.

  • Fei-Fei Li's Startup World Labs is launching its first commercial World Model with Marble, which can be used to create 3D worlds from text and images. In her manifesto, she herself proclaims the AI era of spatial intelligence.

  • Manifest AI challenges the Transformer architecture with Brumby-14B-Base: It does not require an attention mechanism and promises consistent costs even with huge contexts.

Toolbox for the future: New tools and hardware

  • AI helpers for everyday life: Superinbox upgrades your email inbox, Firefox incorporates a privacy-friendly "AI Window," and LinkedIn launches an AI-based people search.

  • Creative and marketing tools: Flowith offers a central workspace for ideas and content, Foreplay optimizes advertising campaigns, and Epidemic Sound "Studio" automatically creates soundtracks for videos.

  • Hardware innovations: The Dobot Rover X1 is coming into your living room as a robot dog, the Stream Ring records voice memos on your finger, and Samsung turns the TV into an assistant with the Vision AI Companion.

  • Autonomous driving: Waymo finally ventures onto the highway, and XPENG presents its vision of the self-driving car as an AI travel companion.

  • AMIGO sharpens blurry images from the James Webb Telescope.

  • The SSi Mantra surgical console enabled the first remote bypass surgery!

That’s a wrap for now! See you next week,
Fabian & Reto

Follow us on Twitter & LinkedIn!

We love hearing from you. Just reply to this email with your feedback—what you liked, what you didn't, or what you're interested in!