|
In this newsletter:
- Google's Gemini Auto Browse handles web tasks inside Chrome
- xAI launches Grok Voice Think Fast 2.0 with faster, smarter conversations
- LinkedIn adds AI slop reporting button and removes its own AI writing tool
Plus, you’ll find new AI tools and this week’s top AI news headlines!
🌐 Google's Gemini Auto Browse Handles Web Tasks Inside Chrome
Gemini Spark is Google's always-on AI agent that works in the background on your behalf - even when your phone is locked or your laptop is closed. It launched at Google I/O in May for Google AI Ultra subscribers and has just received a significant upgrade: it can now browse the web and take action directly in Chrome.
What is Gemini Spark:
- Always-on AI agent running 24/7 on Google's cloud, no need to keep your laptop open
- Receives tasks via a dedicated Gmail address and pulls context from Gmail, Docs, Sheets, and Calendar
- Requires your explicit confirmation before taking any sensitive action
What's new:
- Rolling out now in the US, more countries coming
- Google AI Pro subscribers in over 160 additional countries now have access to Spark
- Spark can use your logged-in Chrome accounts and saved passwords to handle web errands
- Example tasks: scheduling apartment viewings, researching flights, and starting the booking process
- Hands sensitive actions like payments back to you rather than completing them automatically
- Protected against prompt injection attacks
Why it matters:
Most AI assistants stop at telling you what to do next. Spark logging into websites on your behalf and actually doing it is a different category of useful - and the Chrome integration is what makes that possible at scale.
The question now is whether people trust Google enough with their saved passwords to hand over that level of access, which is less a product problem and more a reputation one.
👀 Read more about Gemini Spark's Chrome update!
🗣️ xAI Launches Grok Voice Think Fast 2.0 With Smarter Conversations and Faster Responses
SpaceXAI released Grok Voice Think Fast 2.0, its most capable speech-to-speech voice model, with improved transcription accuracy, faster response times, and better conversational ability across 24 languages.
What's new:
- Available now at $0.08 per minute of audio
- On August 5, 2026, grok-voice-latest automatically upgrades to 2.0; no action needed
- To stay on version 1.0, pin grok-voice-think-fast-1.0 before August 5
- Reasons through queries while speaking, so tool calls usually execute before the end of the first sentence
- Uses 60% fewer reasoning tokens than version 1.0 with no impact on latency
- Speaks in shorter sentences, asks one question at a time, avoids filler
- Tested on Starlink customer support line, showed significant increase in sales conversion and support resolution rates
Performance vs competitors:
- AA Speech-to-Speech Quality Index: 82.9% vs GPT-Realtime-2.1 (79.1%) and Gemini 3.1 Flash (69.5%)
- Conversational Dynamics: 95.1% vs version 1.0 (77.8%) and Gemini 3.1 Flash (74.3%)
- Agentic Performance: 56.5% vs GPT-Realtime-2.1 (45.7%) and Gemini 3.1 Flash (37.7%)
- Time to first audio: 0.70 seconds vs Gemini 3.1 Flash (2.98 seconds)
Transcription accuracy:
- 1.4x improvement over version 1.0
- 1.5 to 2.0x improvement over Deepgram Nova 3 and ElevenLabs Scribe v2
- Gap widens to roughly 10x better than dedicated speech-to-text models in noisy settings with background noise and telephony compression
- Covers 24 languages including English, Arabic, Chinese, Hindi, Japanese, French, German, Spanish, Portuguese, and more
Why it matters:
Voice AI that works well in a quiet demo but falls apart in a noisy call center or car has been the reason most companies haven't replaced human agents yet.
A 10x accuracy gap over dedicated transcription tools in real-world noise removes that excuse, and the Starlink sales conversion improvement shows this isn't just a benchmark story - it's already changing business outcomes in production.
👀 Read more about Grok Voice Think Fast 2.0!
🚫 LinkedIn Lets Users Flag AI-Generated Posts and Removes Its Own AI Writing Tool
LinkedIn added a "Seems like AI slop" button that lets users report low-quality AI-generated content, while also pulling its own "enhance your post" feature, which used AI to rewrite posts in your voice.
What's new:
- New "Seems like AI slop" button on posts to report low-quality AI-generated content
- User reports feed directly into LinkedIn's AI models to help identify slop more accurately
- Blocking hundreds of thousands of automated comment attempts daily and millions of other automation attempts in the past couple of months
- New classifiers to identify AI slop and reduce it from suggested content recommendations outside your network
- Private dashboard flags for users whose posts are coming across as inauthentic due to heavy AI use
- Removing "enhance your post" AI writing feature, replacing it with a proofreader that corrects your words without changing your voice
- Expanding profile and page verification tools
- Adding option to block comments from company pages you no longer want to see
Industry context:
- Substack added a tool last week to identify AI-written content through a partnership with Pangram
- Pangram raised $9 million this week to tackle AI content flooding the internet
- Digg shut down its Reddit competitor in March, overwhelmed by bots
- Cloudflare reports bot traffic now exceeds human-generated web requests
Why it matters:
LinkedIn is shifting from helping users write with AI to helping them sound more authentic. As AI-generated content continues to flood social platforms, success may depend less on using AI and more on demonstrating genuine expertise and original thinking.
👀 Read more about LinkedIn's fight against AI slop!
My Latest LinkedIn & X/Twitter Posts:
- Guide for using Perplexity AI effectively (view post)
- 8 ChatGPT prompts for decision-making and strategic thinking (view post)
- One ChatGPT prompt to build an entire agent in minutes (view post)
- OpenAI 251 free AI prompts to automate creative projects (view post)
- 10-step roadmap to build AI agents from scratch (view post)
In partnership with Kit:
As a creator, your time should be spent doing what you love, not juggling a dozen tools just to run your business.
Kit gives you everything you need in one place:
✅ Build and grow your email list (I use Kit for my newsletter)
✅ Easily monetize with paid newsletters and digital products
✅ Automate your emails with triggers and custom workflows
✅ Track what’s working and optimize with powerful insights
It’s not just email. It’s your entire creator business, simplified.
Join a thriving community of successful creators.
Use Kit for Free — and Start Building Smarter
New AI Tools to Boost Your Productivity:
- VidLux: Turns images into videos.
- UGCVideo: Creates AI UGC videos and ads.
- Seedeo: Edits videos with text prompts.
- Freebeat: Creates music and dance videos.
- RespondThat: Automates email replies and phone calls.
- Thrum: Creates personalized songs.
- ReqBrief: Turns client conversations into project briefs.
- Mict: Creates videos from prompts and images.
- PixCosmos: Creates and edits images with AI.
- Kineo: Creates consistent AI Shorts series.
- Spatius: Creates voice-driven language tutors.
- Text2Everything: Creates diagrams from prompts or chat.
- Listen AI Pro: Records, transcribes, and summarizes conversations.
This Week's Top AI News Headlines:
- OpenAI Introduces GPT-5.6, a Smaller AI Model that Delivers GPT-5-Level Performance at Significantly Lower Cost and Faster Speeds, Making Advanced AI More Affordable for Developers (View Article)
- Anthropic Introduces new Cybersecurity Evaluations that Test whether Claude Can Safely Help Security Teams Investigate Cyberattacks without Enabling Malicious Hackers (View Article)
- Perplexity Renames Spaces to Projects, Expanding its AI Workspace for Organizing Research, Files, and Conversations into Collaborative, Long-Term Projects (View Article)
- Google Brings Nano Banana AI Image Generation to Google Earth, Letting Users Create Photorealistic Visualizations of Places and Landscapes Directly within the Mapping Platform (View Article)
- Google Offers Gemini Free Users Up to 10 AI Video Generations with Gemini Omni During a Limited-Time August 2026 Promotion (View Article)
- Google Launches Lyria 3.5, its Latest AI Music Generation Model with Improved Audio Quality, Better Prompt Following, and More Creative Control for Musicians and Content Creators (View Article)
- Google Brings More Natural Voice Conversations to the Gemini App on macOS, Letting Users Speak with AI More Fluidly through Improved Real-Time Voice Interactions (View Article)
- OpenAI Launches ChatGPT for Academic Researchers, Adding New Tools that Help Scientists Search Literature, Analyze Research, Synthesize Evidence, and Accelerate the Research Process (View Article)
- Apple CEO Tim Cook Signals that AI Features Could Become Part of iCloud+, Hinting at a Future Subscription Strategy for Apple Intelligence Services (View Article)
- Microsoft Introduces Phi-5.6, a Small AI Reasoning Model that Delivers Frontier-Level Performance at Lower Cost and Latency by Optimizing Inference Efficiency Instead of Increasing Model Size (View Article)
- OpenAI Reveals that Simple Inference-Time Configuration Changes Tripled GPT-5's ARC-AGI-3 Benchmark Scores without Retraining the Model, Highlighting the Impact of Optimizing AI Reasoning Settings (View Article)
- Ideogram Launches Character References, Letting Users Generate AI Images with Consistent Characters Across Multiple Scenes and Designs using Reference Images (View Post)
- Google DeepMind Introduces Gemini Robotics ER 2, an AI Model that Gives Robots Stronger Spatial Reasoning and Planning Skills to Help them Understand, Navigate, and Complete Complex Real-World Tasks More Reliably (View Article)
- Friend, the AI Companion Wearable, Returns with a New Voice Assistant, Updated Hardware, and a Significantly Higher Price as the Startup Tries to Revive Interest in its Always-On AI Device (View Article)
- Thinking Machines Lab Introduces Inkling Small, a Compact AI Reasoning Model Designed to Deliver Strong Performance on Everyday Tasks while Running Efficiently on Consumer Hardware (View Article)
- Meta's Oversight Board Expands its Focus Beyond Facebook and Instagram, Setting its Sights on AI Chatbots such as ChatGPT and Claude as Questions Grow Over AI Governance and Content Moderation (View Article)
- Reddit Reports Strong Quarterly Growth but Warns that AI-Powered Search is Changing How People Discover Online Content, Creating New Challenges for User Traffic and Engagement (View Article)
Work With Me:
If you enjoy this newsletter, please forward it to your friends and colleagues.
Follow me on LinkedIn and X/Twitter to see my latest posts.
Have a wonderful week!
Andrew Bolis
|