Today's AI Pulse
Google sues Chinese cybercrime network for exploiting Gemini AI in massive scam campaign
Google has filed a lawsuit to dismantle the Outsider Enterprise network that used its Gemini AI to automate scam text messages. The operation ran 9,000 fake websites, registered 1 million fraudulent domains and sent 2.5 million texts in two weeks, scamming hundreds of thousands and causing multi‑million‑dollar losses.
Also developing:
Using AI to Automate Your Social Media and Your Marketing
In this episode, host explores how AI can automate repetitive marketing and social‑media tasks, freeing entrepreneurs to focus on strategy and growth. Guest Brittany Long, an AI coach, shares practical tools and workflows for content creation, scheduling, and analytics, emphasizing the importance of aligning automation with brand voice. Listeners gain actionable steps to integrate AI into their marketing stack and learn how to balance automation with human creativity.
5 Ways AI Is Supercharging DDoS Attacks
NETSCOUT reported a record 8 million DDoS attacks in the first half of 2025, with AI and automation increasingly powering the surge. AI‑driven DDoS‑for‑hire services now let even novices launch attacks by describing objectives to chatbots, automating attack scheduling, vector selection...
VITRA Teaches Robots Hand Skills Through Egocentric Videos
Why do I say that X is the best learning system the world has ever invented? This is a very technical paper on how world models, er VLAs, make robots a lot smarter. Click the Grok button. And ask: “Hey Grok, can...
v0.23.0 CLI Boosts Usability with Tables, Pinned Todos
v0.23.0 cli changelog 🚢 New features - Markdown table rendering - CLI now properly renders markdown tables for better display of structured data and documentation - Pinned todo plans - Pin important todo items to keep them visible and track tasks during long-running...
Challenges HR Leaders Face in Adopting Enterprise AI - with Raúl Monroig of Bristol Myers Squibb
In this episode, Raúl Monroig, VP of People Organization for Bristol Myers Squibb’s Intercontinental region, discusses how global HR teams can harness enterprise AI to address workforce development challenges at scale. He highlights the rapid acceleration of AI adoption and argues...
AI Supercycle Stalls Without Advanced Network Infrastructure
Thank you for reading my latest article Why The AI Supercycle Will Fail Without Advanced Networks . Here at LinkedIn and at Forbes I regularly write about management and technology trends. To read my future articles simply join my network...
AI-Powered Robot Vacuum Redefines Consumer Gadget Excellence
My floor-cleaning robot has an NVIDIA Jetson card inside. This is how it sees, navigates, and takes action. It is an AI Agent. And it is one of the best consumer electronics gadgets I have ever purchased. Been to the CES show...
From Rabbit Holes to Recommendations: Reddit’s Vishal Gupta
In this episode, Vishal Gupta, Reddit’s engineering manager for machine learning, explains how the platform’s recommender systems have evolved from simple collaborative filtering to deep‑learning and large‑language‑model‑driven multimodal models that power both user feeds and ad relevance. He discusses the...
Google's New 50‑Page Guide Becomes AI Agent Blueprint
If you’re building agentic systems, this is your new bible 🔥 Google just dropped a 50-page guide to real-world AI agents: ↳ Core agent architecture ↳ LLM “brains” ↳ Tools and orchestration ↳ Evaluation and deployment ↳ Self-evolving agents (AlphaEvolve) Link in 🧵 ↓ https://t.co/lZyTjRfYUM
Google's AI Roadmap: Future Models & Agentic Systems
Going live at 1800 CET / 2230 IST / 0900 PST with Logan Kilpatrick to discuss the state of AI, what’s next for developers, and how Google is thinking about future models & agentic systems. Hoping for a focused, insight-heavy conversation. Join...
What Trader Joe Can Teach Us About AI Use Cases
The episode draws a parallel between Trader Joe’s historic "Four Tests" for product selection and a new "Five Tests" framework for evaluating enterprise AI use cases, presented by Michele Pelino. Pelino’s five tenets—high business value, rapid learning, adequate data, defensible advantage, and...
China Takes Lead in Open-Source AI Development
China is clearly leading open-source AI now https://t.co/8bZJbQ5rJM
GTC DC '25 Pregame - Chapter 5: AI for Robotics and Manufacturing
The episode explores how AI is bridging the gap between digital intelligence and physical action in robotics and manufacturing, highlighting NVIDIA’s role in enabling real‑time perception, simulation, and generative design for production lines. Industry pioneers discuss the integration of large...
GTC DC '25 Pregame - Chapter 4: AI for Science
In this 34‑minute pre‑game episode, NVIDIA’s GTC DC team explores how artificial intelligence is reshaping scientific research, highlighting its role as a core discovery tool in labs worldwide. They discuss concrete examples of AI accelerating breakthroughs in fields such as...
GTC DC '25 Pregame - Chapter 3: AI Infrastructure Ecosystem
In this 34‑minute pre‑game episode, NVIDIA’s GTC DC series spotlights the AI infrastructure ecosystem, featuring leaders from energy, data‑center, and power‑system sectors who explain how the physical backbone—servers, cooling, electricity, and partner networks—enables today’s AI breakthroughs. The discussion highlights emerging...
GTC DC '25 Pregame - Chapter 2: Agentic AI for Every Industry
The episode explores how agentic AI—intelligent systems that can plan, reason, and act—is transitioning from research labs to practical applications across diverse industries. Experts discuss real‑world use cases such as autonomous manufacturing optimization, AI‑driven drug discovery, and dynamic supply‑chain management,...
GTC DC '25 Pregame - Chapter 1: State of AI Innovation
The episode surveys the current landscape of AI innovation, highlighting how emerging models, research breakthroughs, and open‑source collaborations are reshaping the field. Investors and startup founders discuss where the next durable wave of AI growth will originate, emphasizing sectors such...
👾 Anthropic Chose Boring Over Viral, and It’s Paying Off
Anthropic deliberately shuns consumer‑focused, viral AI demos in favor of building safe, reliable solutions for heavily regulated enterprises such as banks, hospitals, and Fortune 500 firms. By positioning itself as an unconflicted partner, offering safety‑first models like Claude Code, and scaling through...
Google Will Peel Back a New Era of AI Images with Nano Banana 2
Google previewed Nano Banana 2, the next iteration of its AI image generator, inside the Gemini app. The model introduces a multi‑step workflow that plans an image, detects errors, and iteratively corrects them, delivering sharper angles, more accurate colors and editable text...
Lovable Says It’s Nearing 8 Million Users as the Year-Old AI Coding Startup Eyes More Corporate Employees
Lovable, the Stockholm‑based AI coding platform founded a year ago, is approaching 8 million users, up from 2.3 million in July, and reports 100,000 new products built daily. The company, which raised $228 million—including a $200 million Series round that valued it at $1.8 billion—hit...
Hiring AI‑Tool Tester & Writer for Growing Platform
We're hiring someone to get paid to play with AI tools all day. (Not kidding!) The Rundown is growing fast (2M+ readers), and we're looking for another writer who's obsessed with AI to join us. Your job: > Test every new AI tool that...
⚡ Inside GitHub’s AI Revolution: Jared Palmer Reveals Agent HQ & The Future of Coding Agents
In this episode, Jared Palmer—GitHub’s SVP and Microsoft’s VP of CoreAI—discusses the rapid evolution of coding agents, the launch of Agent HQ as a collaborative hub, and the breakthrough Next.js coding agent v0 that emerged from tight platform constraints. He...
Study Finds AI Models Store Memories and Logic in Different Neural Regions
Researchers at Goodfire.ai demonstrated that memorization and logical reasoning in large language models occupy distinct neural pathways. By surgically pruning low‑curvature weight components, they eliminated 97% of verbatim recall while preserving 95‑106% of reasoning performance on benchmarks such as BoolQ...
Apple's Rumored 'Health+' Subscription Tier with AI Features Is Doomed to Repeat Garmin's Failure
Apple is preparing to launch a paid Health+ tier that will bundle AI‑driven nutrition and medical advice into the Health app, accessible on iPhone and iPad rather than the Watch. The move mirrors Garmin’s recent Connect+ subscription, which sparked user...
Here’s How to Get ChatGPT to Finally Understand You
PromptBuilder is marketing a lifetime AI prompt‑engineering tool for $199, a steep discount from its regular $1,764 price. The software lets users select templates, input brand‑specific details and automatically generate optimized prompts that work across ChatGPT, Claude, Gemini, DALL‑E and...
Siri Shortcuts Now Works with Samsung SmartThings Routines
Samsung’s latest SmartThings update adds Siri Shortcuts support, letting iPhone, HomePod and other Apple devices trigger SmartThings routines via voice. Users can now ask Siri to run automations such as a “Good morning” routine that adjusts lights, starts the coffee...
Can AI Help Nonprofits Do More with Less?
A 2025 Candid report found that 65% of U.S. nonprofits are interested in AI, though most are only at a beginner level, and a Bonterra survey shows over half of its partner nonprofits have already deployed AI, especially small organizations...
Paramount Plus Is Getting a Price Hike in Early 2026
Paramount Plus will raise its U.S. subscription fees on Jan. 15, 2026, moving the ad‑supported Essential plan from $7.99 to $8.99 and the ad‑free Premium plan from $12.99 to $13.99. The hike follows a series of annual price increases and comes after the...
AI Now Handles Multi-Day Projects Without Human Supervision
Ran into a CEO in the park and showed them something that changed how they're thinking about the future of AI. I think you might want to hear this. Claude Sonnet 4.5 completed a 30-hour coding task with zero human...
A Better Way of Thinking About the AI Bubble
Russell Brandom argues that the AI "bubble" is less about a sudden crash and more about a massive, mismatched bet between fast‑moving AI software and the slow‑to‑build data‑center supply chain. He notes that billions of dollars are already pledged—Oracle’s $18 billion...
Google AI Studio Scores 8/10 for Free, Seamless Integration
Google AI Studio Build Mode Review - I give it an 8/10. - comparable to other AI app builders in quality, but free to use (they must be losing money on this, no?) - turns simple/broad ideas into working micro apps in...
What Should the Government’s Role in AI Be?
The episode examines the evolving debate over government involvement in AI, sparked by OpenAI’s recent "backstop" comments and its AI Progress and Recommendations report. It breaks down arguments for and against bailouts, industrial policy, and compute subsidies, featuring perspectives from...
AI Creativity Turns Work Into Playful Future
I have been talking to the guy (and it is a guy who lives in the UK) who is behind @trishacode (his wife is the drummer). He is one of a new crop of creative geniuses using AI to create new...
Gemini for TV Is Coming to Google TV Streamer Starting Today
Google is rolling out Gemini for TV to Google TV Streamer devices starting today, extending the generative‑AI assistant beyond the limited set of smart TVs that received it in September. The update works on any TV with an HDMI port,...
Google Brings Gemini to the Google TV Streamer
Google announced it will roll out its Gemini AI to the Google TV Streamer, replacing Google Assistant and enabling more natural, conversational voice interactions for content discovery and general queries. Users can ask nuanced questions like movie recommendations for mixed...
Chronosphere Takes on Datadog with AI that Explains Itself, Not Just Outages
Chronosphere, the $1.6 billion New‑York observability startup, announced AI‑Guided Troubleshooting built around a Temporal Knowledge Graph that continuously maps services, dependencies and change events. The feature delivers data‑backed “Suggestions,” investigation notebooks and natural‑language queries, keeping engineers in control by showing the...
Wikipedia Urges AI Companies to Use Its Paid API, and Stop Scraping
Wikipedia’s nonprofit Wikimedia Foundation announced that AI developers must access its content via the paid Wikimedia Enterprise API rather than scraping, aiming to reduce server strain and generate revenue for its mission. The blog post highlighted recent spikes in AI...
Deep Learning Demystified: Intuition‑Driven Modern Stack Guide
Deep learning is not a collection of black-box tricks, contrary to what many believe. It can be learned as a principled engineering discipline. This latest edition of Deep Learning with Python is my best attempt so far at teaching it. It...
Meta Launches ASR for 1,600 Languages, Including 500 New
Introducing Meta Omnilingual Automatic Speech Recognition (ASR), a suite of models providing ASR capabilities for over 1,600 languages, including 500 low-coverage languages never before served by any ASR system. While most ASR systems focus on a limited set of languages that...
Forget the Robots – This Is the Reason AI Is the Best Thing to Happen Since Smartphone Cameras
Google has integrated its Gemini generative AI model into Google Photos via an "Ask" button, allowing users to query their personal photo libraries with natural language. In a test case, the author, with a 172,000‑image archive, typed "Picture of bay...
IRobot’s Revenue Has Tanked and It’s Almost Out of Cash
iRobot reported a third‑quarter 2025 revenue shortfall and disclosed that cash reserves have fallen below $25 million, leaving the company without immediate sources of additional capital. The decline follows years of competitive pressure from Chinese rivals, a failed Amazon acquisition, large...
Open‑source dLLM Library Brings Diffusion LLMs Near Llama 3 Performance
Wow. What an ace (free!) library for building Diffusion Language Models! 🤗 dLLM is a new open-source framework that unifies training, inference, and evaluation for diffusion-based LLMs. Unlike autoregressive models that generate text token by token, diffusion models rely on an iterative denoising...
Investing in Tiny AI Chips for Robots, Medical Devices
This company is one you will hear a LOT more about. They make tiny AI computers for robots and medical devices. I just invested after a friend who makes medical devices told me how far ahead it is.
World Models Soon Control Everything— We’re Ignoring It
World models are soon going to run everything. The world isn’t paying enough attention.
VCs Tout Elite AI Labs; Founders Use Alibaba's Qwen
venture capitalists: we back the top ai labs and infrastructure companies such as openai anthropic mistral huggingface coreweave cohere xAi ai founders: we use qwen on ali baba cloud
One-Click Access to Any AI Model, Built for You
Every AI model in the world—open source, commercial, audio, image, text—under one-click. Just ask Agent to build with your favorite model and let us worry about the rest.
Meet the AI Realist: Optimistic Yet Cautious
What is the official term for someone who is neither an AI Doomer not an AI Boomer? Someone who sees the enormous potential of AI to do good and create abundance, but is also realistic and cautious about the current...
New AI Advances Promise Safe Humanoid Robots Soon
Luckily new AIs are on their way. Active inferencing and cognitive architectures. That will lead to safe humanoids soon.
Meta's AI Glasses Fall Short versus Superior Chinese Rivals
This is why I was so disappointed in Meta’s new display AI glasses. It had one shot to compete with the Chinese and it fell flat. These are way better.
50 AI Tools Driving 24 B Visits Worldwide
Discover the top 50 AI tools that have garnered over 24B visits! These powerful tools are revolutionizing the way we work and interact with technology. Stay ahead of the curve and explore the cutting-edge advancements in artificial intelligence. Via @ingliguori #AI...