- The Rundown AI
- Posts
- Chrome just got 3 new AI features
Chrome just got 3 new AI features
PLUS: Google cuts ties with AI training contractor
Rowan Cheung
January 24, 2024
Welcome, AI enthusiasts.
Google Chrome just got 3 new AI upgrades to help boost user productivity.
The latest features are integrating the tech even further into browsing habits — slowly nudging users towards the AI-driven future of the internet. Let’s explore…
In today’s AI rundown:
New Chrome update brings AI boosts
Google cuts ties with AI contractor Appen
How to build GPTs with better memory
ActAnywhere places subjects into custom video scenes
6 new AI tools & 4 new AI jobs
More AI & tech news
Read time: 4 minutes
LATEST DEVELOPMENTS
Image source: Google
The Rundown: Google just pushed out its latest Chrome browser update, introducing new experimental AI features to improve web browsing – including automatic tab grouping, custom theme generation, and writing assistance.
The details:
Tab Organizer automatically sorts open tabs into suggested groups, making it easier to manage multiple tasks.
AI themes will allow users to personalize their browser by selecting a subject, mood, style, and color for the AI to render.
A new "Help me write" feature launching next month will provide AI-generated suggestions when composing text on websites.
Why it matters: Google continues to make subtle AI integrations into its existing product base — and while these features aren’t going to blow anyone away, slowly shifting consumer habits towards using the advancing tech is a likely win in the long run.
TOGETHER WITH TÜV SÜD
The Rundowns: The AI Quality Framework helps manage the risks of regulation, security, privacy and ethics for a successful AI integration.
TÜV SÜD offers:
Training: Customised training for executives and AI specialists.
Risk Profiling: Systematic identification and assessment of potential risks to ensure responsible deployment and management.
Regulatory Compliance: Assistance with the EU AI Act and other national regulations
Leverage our AI expertise for compliance, ethics, and scalable solutions. Visit us to start your journey in AI excellence.
Image source: CNBC
The Rundown: Google just terminated its contract with Appen, an Australian firm that provided training data to improve AI systems like Bard — a move that the Alphabet Worker’s Union says will have a ‘devastating impact’ on its subcontracted employees.
The details:
Appen employs over 1M contractors globally to label data for training AI models, touting Microsoft, Apple, Meta, and Amazon on their client list.
The contract with Google accounted for nearly a third of Appen's revenue — with the company’s shares dropping 40% on the news.
Google’s employee union said in a statement that the move will impact ‘at least 2000’ subcontracted workers.
Appen has faced criticism in the past over low wages for contractors across the globe working on Google products.
Why it matters: Google’s wave of layoffs and restructuring isn’t limited to its full time staff — with partners and contractors now feeling the heat as well. But as AI shifts from human labor to more compute power, these ‘ghost workers’ may quickly find themselves obsolete in the training process.
AI TRAINING
Nick Dobos (@NickADobos) on X
The Rundown: Loading up your GPT with knowledge files may seem logical, but it can tank performance by flooding the context window. Follow these best practices below to improve your bot’s memory.
Thank you to Nick Dobos (a top 10 GPT creator) for sharing these tips.
Recommendations to improve GPT performance:
Instead of inserting documents for knowledge files, use .md (you can save files in this format by renaming them in a code editor).
This workaround keeps information out of the context window, giving the GPT more available memory.
Only fetch files when needed with Code Interpreter.
Do not use all 20 files if possible — but if unavoidable, keep them small (.md format helps this).
Only include actionable docs with working system prompts for the AI.
AI RESEARCH
Image source: Adobe
The Rundown: Adobe researchers just developed an AI model called ActAnywhere, which can digitally insert a person into new video backgrounds — creating realistic footage of them moving and interacting within the scene.
The details:
ActAnywhere takes video input of a person and a still image defining the new background.
Using deep learning, it generates a video showing the person embedded into the scene with proper motion, scale, lighting, and shadows.
In tests, ActAnywhere produced highly realistic videos — with water splashing as a subject runs through puddles or smoke wafting from fog.
The model automates the tedious production process of realistically blending a subject into a new background.
Why it matters: ActAnywhere’s impressive breakthrough could have major professional applications in filmmaking, video effects, content creation, and beyond — saving production time while still generating hyper-realistic outputs.
NEW TOOLS & JOBS
🗣️ Deepgram- Introducing Aura, a powerful text-to-speech (TTS) API built for real-time conversations. Be the first to try Aura by joining the waitlist (link)
🤖 Top GPTs- Access the best OpenAI GPTs without a Plus subscription (link)
👨💼 KPI Builder- Find the KPIs you should care most about as a founder (link)
💻 Brainner- Streamline talent acquisition by automating AI-driven resume analysis (link)
🎨 Recraft- Create and edit graphics in a uniform brand style (link)
👨💻 brikly- Platform for recruiting engineers on an agency basis (link)
QUICK HITS
AI researcher Jim Fan gave a Ted Talk presentation unveiling the concept of a ‘foundation agent’ as the next challenge for AI — envisioning a model that operates across both virtual and physical worlds.
AiDash announced a $50M raise for its AI systems that combine with satellites to detect wildfire and weather risks near power lines.
Google Ads is integrating the company’s Gemini AI model to help advertisers build better search campaigns, showing a 42% increase in ad strength for small businesses.
The United States Postal Service is leveraging generative AI integrations as it modernizes its services with machine learning across hiring and operations.
Kin.art launched a free tool for artists to prevent their artwork from being used in generative AI model training by subtly altering the images.
The UK Government declared AI's reproduction of copyrighted works illegal without a license or exception, rejecting exceptions for text and data mining.
Planet and Microsoft are collaborating to create an AI-focused resource for environmental reporting, aiming to improve sustainability and biodiversity tracking.
THAT’S A WRAP
SPONSOR US
Get your product in front of over 450k+ AI enthusiasts
Our newsletter is read by thousands of tech professionals, investors, engineers, managers, and business owners around the world. Get in touch today.
FEEDBACK
How would you rate today's newsletter?Vote below to help us improve the newsletter for you. |
If you have specific feedback or anything interesting you’d like to share, please let us know by replying to this email.
Join the conversation