Black Lotus Labs null-routed over 550 C2 nodes associated with the AISURU/Kimwolf botnet, which has infected more than 2 million Android devices—primarily unsanctioned TV streaming boxes with exposed Android Debug Bridge (ADB) services. The botnet exploits vulnerabilities in residential proxy services to propagate, turning compromised devices into proxy nodes for DDoS attacks and malicious traffic relay. Security teams should monitor for ADB exposure on network devices and block traffic to the identified C2 infrastructure.
A security-conscious user received several 2FA prompts and SMS verification codes for their Apple account one afternoon, followed by an automated call warning them that their account was under attack. After receiving this call, a second call came from a very professional-sounding employee who created a support case for the user, verified that it was from Apple to build trust, and walked the user through resetting their password. The phish then had the user “close the case” via an SMS link, which directed them to a sophisticated phishing site where they could steal the user's credentials.
CERT-UA reported that the suspected Russian threat group Void Blizzard targeted Ukrainian Defense Forces officials via Signal and WhatsApp with charity-themed lures delivering the PluggyApe backdoor, which now features MQTT-based C2 communication and dynamic address retrieval from paste sites.
The WEF's fifth annual cybersecurity report shows 94% see AI as the main cyber driver, 87% cite AI vulnerabilities as the fastest-growing risk, and 91% of large organizations are revising cybersecurity strategies due to geopolitical volatility.
Anthropic is donating $1.5 million to the Python Software Foundation to enhance security in CPython and PyPI, thereby lowering supply chain risks for millions of Python developers and possibly other open source ecosystems too.
Yes, you can manage creative campaigns without drowning in your inbox. When your team can see every deadline, asset, and stakeholder comment in one place, the “just following up” emails evaporate.
Google traffic to publishers fell sharply in 2025 and is expected to keep dropping through 2026. Search fell 33%, and Discover fell 21%, with deeper losses since May 2023 and even steeper declines in the US. Lifestyle and utility sites were hit hardest as AI summaries reduced clicks. Publishers expect a 43% drop over three years and plan to scale back on Google, Facebook, and X while shifting more effort to AI platforms and YouTube. Most expect some AI licensing revenue, but not much. Confidence in journalism is falling, and leaders are leaning into creator partnerships and subscriptions to counter shrinking platform visibility.
A simple daily check-in app has gained rapid traction by addressing fears tied to solo living rather than offering companionship. The app requires one tap per day and alerts an emergency contact after two missed days, and it has already charted in markets such as the USA and Australia after rebranding as Demumu. Its rise aligns with structural shifts, as China had more than 125 million one-person households in 2020, and the solo living rate exceeds 30%, with projections reaching 200 million households by 2030. The popularity highlights how basic visibility and being noticed matter to people who feel isolated and lack strong offline support, creating opportunities for brands to design products and experiences for solo dwellers.
Well-attended webinars start with a tactical theme and the right guest, chosen around either a topic or their expertise. A 1:1 brainstorm with the SME generates content, then AI organizes transcripts, social posts, and context into a deck outline to reduce the guest's workload. A detailed run-of-show is created with timing, sections, roles, and talking points, followed by a first draft of the deck. A 30-60 minute session with the guest reviews the run-of-show and deck, gathers feedback, and finalizes the deck as a playbook. During the event, engagement is actively facilitated. Then immediate follow-up provides a post-event deal room with recordings, slides, resources, and speaker info.
Google AI Overviews frequently summarize Reddit and Quora discussions as authoritative answers, which allows outdated complaints or fringe opinions to appear factual. Brands should continuously monitor high-engagement forum threads and publish authoritative, AI-readable content that clearly states verified information and corrections. The goal is to replace forum consensus with primary sources that AI systems can confidently cite.
Claude Cowork is a research preview for Claude Max subscribers on macOS that lets anyone (not just developers) work with files in a chosen folder. Claude can read, edit, and create documents, spreadsheets, or reports, plan tasks, and work in parallel while keeping users informed. It can connect to external tools and handle tasks online with browser access. Users control which folders and connectors Claude can access, and it asks before major actions to prevent mistakes.
Google updated Veo 3.1 to generate short-form video using up to 3 reference images so visual elements remain stable across outputs instead of drifting between generations. The update enables native 9:16 vertical video, making clips immediately usable for social-first distribution.
Using hand gestures while presenting increases perceived competence. It also increases persuasiveness and positive audience reactions by up to 9%. In an analysis of 2,184 TED talks and 200,000 clips, each doubling of hand movements led to 5.18% more YouTube likes. In experiments, a circular gesture during a face cream pitch drove 9% more positive responses than no motion and 7.7% more than pointing.
Generative AI is weakening the discovery advantage that online travel platforms like Expedia and Booking.com built through search-driven aggregation. As consumers plan trips inside conversational AI tools, intent is forming outside aggregator websites, reducing their ability to influence decisions and monetize early-stage demand. To defend their position, incumbents are integrating with AI platforms like Google and ChatGPT while experimenting with generative features inside their own apps.
Canva reached product-market fit by moving prototyping relentlessly with non-designers and letting real user friction drive product decisions rather than investor pressure or launch theatrics. Early research revealed that users froze when faced with a blank design surface, which led to a short onboarding video and guided interaction that improved activation and retention. From there, growth compounded through early evangelists like social media managers, programmatic SEO that dropped users directly into creation, aggressive global localization, and disciplined technical and financial decisions that enabled long-term scale.
Instagram's “Your Algorithm,” available to English-speaking users, allows people to directly influence their Reels feed by selecting and adjusting topic preferences.
Apple recently announced that it had decided to rely on Gemini to power its AI features. The company will be able to fine-tune the model so that it responds in Apple's preferred manner. Apple's AI won't respond with any branding related to Google or Gemini. Some of the features will launch this spring. More advanced features are expected to be announced at Apple's annual developer conference in June.
Meta and EssilorLuxottica are discussing potentially doubling production capacity to 20 million units or more by the end of the year to capture growing demand and head off rivals. They may potentially further establish the capability of producing more than 30 million units if demand justifies such a move. No final decisions have been made yet. The talks signal confidence that smart glasses can move beyond early adopters and reach mass-market scale.
China's Institute of Radio Spectrum Utilization and Technological Innovation has filed a request with the International Telecommunications Union to operate a satellite constellation with a total of 193,428 satellites. The request far exceeds the ambitions of other current constellation projects. The Chinese constellations plan to use a wide range of radio frequencies and will be deployed across numerous orbits from 300 to 600 kilometers, as well as higher objects at around 20,000 kilometers. The US Federal Communications Commission views China's satellite ambitions as a national security threat.
Reports from a supply chain leaker indicate that OpenAI's first AI gadget is aimed at taking on Apple's AirPods. The device, codenamed 'Sweetpea', is an audio product that resembles an earpiece. It is projected to be released around September. OpenAI has other hardware in the works, including a home-style device and a pen.
Ramp [8] adopted the Builder mindset and an ambitious plan: to become the most productive company in the world. So far, they shipped 270 features in H1 2025 (MORE THAN ALL OF 2024 combined), and 90% of 1,200 employees use Notion AI monthly. Read more on the Notion blog [8] →
Smart specs guide agents clearly, stay within practical context sizes, and evolve with projects. Aim to create clear specs that cover just enough nuance to guide the AI without overwhelming it. Break large tasks into smaller ones. Plan first in read-only mode before executing, and iterate continuously. This post presents a framework for spec-writing that keeps AI agents focused and productive.
Google deprecated JPEG-XL support in Chrome/Chromium back in 2022. The decision was widely criticized, and Google said that it may reconsider. The code to restore support was merged this week, so JPEG-XL image support is back in the very latest Google Chrome/Chromium codebase.
AI should be part of every process, but the final filter must be human judgment informed by taste. The hardest decisions aren't reducible to rules. AI can tell you whether something works, but only taste can tell you whether it belongs. Personal taste is the moat when execution is cheap and correctness is abundant.
The tech world is moving towards voice-powered computing. Several companies are building purpose-built hardware for AI that will be predominantly controlled via voice. This will help set up the next wave of technology: robots. Robots will be wholly voice-controlled.
The partnership gives Apple time to manage expectations across its platforms as it develops its own in-house AI server chips, expected to enter mass production in the second half of 2026.
Today's masses still have power due to their labor contribution and their ability to physically rebel against the elite, but AI has the potential to change that power balance.
LTX-2 is officially out, and you need to see it to believe it [4]. Produce up to 20 seconds of high-fidelity video [4] with complete creative control and consistent style.
Anthropic Labs has expanded its team to incubate experimental products at the forefront of AI capabilities, leveraging rapid model advancements. Co-founder of Instagram, Mike Krieger, has joined Labs for collaborative product development. Ami Vora will lead product organization alongside CTO Rahul Patil to scale Claude experiences. Recent successes include Claude Code's billion-dollar growth and the Model Context Protocol's rise to industry standard.
The Veo 3.1 update introduces new features designed for both casual creators and professionals. Users can now generate videos based on reference images. The tool now supports native vertical outputs tailored for mobile devices, and high-resolution upscales to 1080p and 4K are also available. The release is available now across multiple Google services and platforms.
Google is testing an Auto Browse tool for Gemini that lets the AI manage tasks in Chrome. The tool will allow Gemini to autonomously browse the web, manage tabs, and directly interact with Chrome on users' behalf. Code references indicate that Auto Browse may be exclusive to the Gemini Ultra plan. Google is rolling out the feature incrementally to a limited audience in the US.
Tsinghua University and Zhipu co-hosted AGI-Next, a summit for frontier AI, in Beijing on January 10. The event included a series of keynotes by notable figures in the Chinese AI industry. This post contains a translated transcript of the event, revealing a fascinating conversation about the AI landscape in China that covers the technical side, corporate dynamics, as well as the future envisioned by China's industry Titans. The event took an honest look at whether China's open-source leadership has actually narrowed the technology gap with the US, China's emerging AI-for-business paradigm, and what it will take for Chinese researchers to take riskier bets.
AI-assisted coding streamlines product creation but often sacrifices long-term design sustainability. It prioritizes immediate functionality, leading to hardcoded configurations, which are fragile for future updates. Successful product development requires experienced system design to avoid a "build-now, regret-later" scenario, transforming AI vibes into structured, sustainable solutions.
The AI Chip Sales data explorer uses financial reports, company disclosures, and more to estimate compute, power usage, and spending over time for a wide variety of AI chips. The cumulative global AI compute capacity has reached the equivalent of more than 15 million Nvidia H100 GPUs. The cost to purchase these chips has rapidly escalated to tens of billions of dollars per quarter, drawing over 10 GW of power per quarter. A link to the interactive data explorer is available.
Unblocked [11] is the AI code review that surfaces real issues and meaningful feedback instead of flooding your PRs with stylistic nitpicks and low-value comments.
DeepSeek's Engram architecture introduces a conditional memory system that uses lookup tables for common N-gram patterns, significantly reducing computational overhead in LLMs. This approach boosts performance on knowledge benchmarks and reasoning tasks by efficiently freeing up neural resources for complex reasoning. Engram's design leverages tokenizer compression, multi-head hashing, and context-aware gating, providing a marked improvement over MoE baselines without substantial memory impact.
GLM-Image is an open-source, industrial-grade discrete auto-regressive image generation model with a hybrid architecture that combines an auto-regressive module with a diffusion decoder. It excels in text-rendering and knowledge-intensive generation scenarios. The model performs especially well in tasks that require precise semantic and complex information expression. It supports a rich set of image-to-image tasks.
Vocal computing may finally become mainstream due to new AI advancements in LLMs overcoming past limitations of systems like Siri and Alexa. OpenAI's breakthroughs and its acquisition of a hardware startup hint at innovative voice-driven devices potentially moving beyond the smartphone. As tech companies refine and develop diverse hardware for AI, voice control is pivotal in advancing towards a more integrated and intuitive tech ecosystem.
AI introduces "borrowed certainty" in software, creating technical debt by replacing proof and evidence with confident-seeming predictions. Developers often mistake AI's fluency for validity, risking decisions based on unearned certainty. Best practices include treating AI output as a hypothesis, demanding traceability, ensuring human understanding, and pairing AI with execution feedback.