Search the archive

Get $25 in credit and start building

NVIDIA, OTHERS IN TALKS FOR OPENAI FUNDING, INFORMATION SAYS

Nvidia, Microsoft, and Amazon are in talks to invest as much as $60 million into OpenAI. The total round will contribute as much as $100 billion to OpenAI's funds, which could help alleviate investor concerns about OpenAI's cash burn while continuing the investment frenzy. OpenAI is also attempting to line up funding from top investors in the Middle East in a funding round that could total at least $50 billion. These talks are early, and the amounts could change.

GOOGLE BEGINS ROLLING OUT CHROME'S “AUTO BROWSE” AI AGENT TODAY

Google has started rolling out its autonomous browsing agent to Chrome. The new features are accessible from the AI button, which will now default to a split-screen view. Users can still pop Gemini out into a floating window, but the split-view gives the tool more room to breathe while manipulating a page with AI. Gemini in Chrome can access and edit images with Nano Banana - users just have to open an image from the web and type in the side panel with a description of the edits they want.

META'S AI PLANS INCLUDE SHOPPING AGENTS AND NEW MODELS

Meta announced a major AI product rollout for 2026, emphasizing new models and agentic tools focused on commerce. Mark Zuckerberg pointed to personalized AI shopping assistants and Meta's access to user context as key differentiators in delivering tailored agent experiences.

TRAINING A 67M-PARAMETER TRANSFORMER ON AN M4 MAC MINI

This paper takes a grounded look at what happens when you build a modern language model from first principles, train it on real data, and ask it to do something unforgiving. The author trained a 67-million-parameter transformer end-to-end on an M4 Mac Mini using Apple Silicon MPS - no discrete GPU, just 24 gigabytes of unified memory. The resulting model achieved 93.94% exact-match accuracy on CLI command generation.

ON-DEVICE LLMS IN 2026

Real-time on-device LLMs have become viable due to new deployment techniques and smarter model compression.

RESEARCH AFTER AI: PRINCIPLES FOR ACCELERATED EXPLORATION

AI integration in research transforms it from a tool to an embedded cognitive environment, accelerating exploration but demanding careful application and judgment. Researchers use AI to navigate complex problems, stress-test hypotheses, and facilitate code development and material synthesis, necessitating thoughtful oversight to prevent errors and misinterpretations. While AI speeds exploration, the responsibility for correctness, judgment, and ethical considerations remains with the human authors, addressing challenges of truth, authorship, and ethical implications.

FULLY REMOTE)

As the first engineer on TLDR's new Applied AI team, you'll build AI agents and composable Claude Skills to let non-technical teammates create their own AI workflows. Learn more [13].

TOWARDS A SCIENCE OF SCALING AGENT SYSTEMS: WHEN AND WHY AGENT SYSTEMS WORK

Smarter models accelerate the need for multi-agent systems. However, the 'more agents' approach often hits a ceiling. It can even degrade performance if not aligned with the specific properties of the task. The next generation of AI agents needs to move from heuristics to quantitative principles to become more numerous, smarter, safer, and efficient.

THE SURPRISING ATTENTION ON SPRITES, EXE.DEV, AND SHELLBOX

Several new Linux VPS services have popped up recently, all optimized to get users running from zero to 'machine running Claude Code' in a couple of minutes. They also handle the annoying bits of sharing web services, going beyond just the 'development' part and to the full end-to-end stack. These services are targeting developers who want frictionless access to machines without having to set up VMs, wrestle with containers, or configure certificates to show off prototypes. For many developers, it's nice to have something that just works.

MISTRAL VIBE 2.0

Mistral Vibe 2.0 introduces a revamped terminal-native coding assistant with features like custom subagents, slash-command skills, and workflow configuration. It's powered by the new Devstral 2 model family and available via Le Chat Pro and Team plans.

VISUAL MODELS FOR REASONING LIKE HUMANS

This project introduces a unified multimodal model that uses visual generation as part of chain-of-thought reasoning, aiming to mimic human-like reasoning through both visual and verbal modalities.

YOUTU-VL: A LIGHTWEIGHT VLM

Youtu-VL is a compact 4B-parameter vision-language model that uses a novel training strategy called Vision-Language Unified Autoregressive Supervision (VLUAS). This approach boosts general and vision-centric task performance without relying on task-specific modules.

WHY AI SWARMS CANNOT BUILD ARCHITECTURE

Agents can be individually excellent but collectively incoherent. Two agents with identical training who independently come up with the same answer aren't coordinating, they're sampling from the same distribution. Swarms are parallel, independent, and stateless. They have no propagation mechanism, and scaling makes it worse. Adding global visibility, temporal persistence, and enforcement authority to a swarm would change it into a hierarchical system with stochastic workers, creating an architecture that allows for coordination.

WORLD MODELS

The AI industry is turning its focus to world models, models that predict the next state observations. These models aim to understand the causal laws of an environment like a video game, codebase, or market. Companies have been working on similar models for decades under different names, but the bottleneck to the next step is now almost able to be overcome. The next AI era will require learning from the world directly through execution and outcomes.

MARK ZUCKERBERG SAYS A FUTURE WITHOUT SMART GLASSES IS ‘HARD TO IMAGINE'

Mark Zuckerberg claims AI smart glasses are set to become mainstream.

KEEPING YOUR DATA SAFE WHEN AN AI AGENT CLICKS A LINK

OpenAI's new safety mechanism is designed to reduce the risk of AI agents inadvertently leaking private data when they automatically fetch URLs by only allowing automatic retrieval of links that are publicly indexed on the open web.

GOOGLE TESTS VOICE CLONING ON AI STUDIO POWERED BY GEMINI

Google is testing voice cloning for AI Studio with a hidden "Create Your Voice" feature linked to the Gemini 2.5 Flash model.

NOMURA'S LASER DIGITAL APPLIES FOR US NATIONAL TRUST BANK

Laser Digital, Nomura's digital asset arm, filed for an OCC national trust bank license to launch institutional crypto custody, spot trading, and staking. Joining Ripple and Circle, the firm targets federally regulated infrastructure to meet increasing institutional demand for secure, governed access to digital asset markets.

ALTCOINS JUMP AS DOLLAR SLIDES, BITCOIN HOLDS STEADY

The Dollar Index fell to a four-year low, breaking below a 2011 trendline, which supported crypto prices with bitcoin holding near $89,200 and ether topping $3,000. Altcoins significantly outperformed the majors, with the altcoin-heavy CD80 index posting a 3.7% daily gain versus 2.47% for bitcoin-dominant CD20. The altcoin strength reflects typical rotation patterns as bitcoin remains range-bound, with capital flowing to more speculative assets. Despite the upside, derivatives markets saw another $230 million in long liquidations over 24 hours, while put options continue trading at a premium to calls with increased demand for downside hedges at the $85,000 strike.

PREDICTION CAPITAL MARKETS

Despite generating $150B+ in volume and $800M in revenue through platforms like Pump since January 2024, creator tokens suffer from a fundamental structural flaw: perpetual onchain instruments tracking ephemeral attention create perpetual bagholders by design. The lack of objective resolution mechanisms means winners are determined by exit timing rather than correctness, enabling manipulation while forcing creators to absorb direct reputation damage whenever they withdraw fees or token prices decline. The proposed solution replaces perpetual creator tokens with time-bounded markets featuring defined speculation windows and resolution endpoints that match the ephemeral nature of attention cycles.

EIGENCLOUD: VERIFIABLE OFF-CHAIN COMPUTATION

EigenCloud introduces a hybrid verification model that combines Trusted Execution Environments (TEE) with restaking-based economic security to enable verifiable off-chain computation, addressing the fundamental lack of transparency in traditional cloud services while circumventing blockchain's computational limitations. The platform allows complex operations like AI inference and GPU-intensive workloads to execute off-chain with cryptographic proof of correct execution, bridging the gap between Web2 scalability and Web3 verifiability. Its architecture supports familiar developer tools, including Docker containers and external API calls, requiring minimal smart contract expertise to implement verifiable applications. Target use cases span prediction markets, institutional finance, cross-chain security infrastructure, and AI agent ecosystems where result verification and trustless coordination are critical.

THE 7 POWERS IN CRYPTO

Traditional moats like scale economies, network effects, and switching costs are being eroded by open-source code and composability. Certain advantages survive: "Lindyness" (Aave capturing 100x more lending supply than competitors at lower rates due to multi-cycle security track record), liquidity network effects (Pump's bonding curve mechanizing tipping points), counter-positioning (Morpho's isolated markets vs Aave's monolithic approach), and brand power (CryptoPunks' cryptographically-enforced provenance). Durable crypto advantage accrues around mechanisms that make ownership credibly scarce – "internet bearer assets" – where security history becomes a competitive moat.

WHEN NETWORK BECOMES THE MEDIUM

Blockchain-based art transcends traditional preservation by embedding creation directly into network infrastructure. CryptoPunks has generated $3.07B in trading volume and achieved institutional recognition from MoMA, Centre Pompidou, LACMA, and others, positioning creators Matt Hall and John Watkinson among top-selling living artists. Networked art maintained by global consensus may outlive traditional museums. The Infinite NODE Foundation is launching a Larva Labs exhibition featuring a live Ethereum node sculpture that visualizes real-time network activity.

EUROPEAN STABLECOINS ARE DIVERSIFYING BEYOND THE DOLLAR

EUR stablecoins reached a $617M market cap in early 2026 (Circle's EURC alone at $398M), Swiss CHF stablecoins hit $17.7M, overtaking GBP's $6.1M, and UAH stablecoins total $1.2M with strong local usage. Europe's stablecoin market is diversifying beyond USD dominance with clear upside for payments-driven GBP and regional coins.

FOLLOW THE FEES: STABLECOIN ISSUERS ARE THE TOLL BOOTH

A fee map shows stablecoin issuers dominating DeFi's cashflow. Tether $417.9M (30D) and Circle $196.2M (30D) lead the leaderboard while DeFi-native issuers Sky ($33.1M, +31.7%) and Ethena ($23.3M, +14.9%) are the fastest growers, illustrating that whoever controls on-chain USD rails captures persistent, pump-proof revenue from settlement, lending, margin, liquidity, and treasury flows.

HOW POLKADOT BECAME PROFITABLE FOR THE FIRST TIME IN THREE YEARS

Polkadot secured a 4.1 million dollar profit in Q4 2025.

CHINESE MONEY LAUNDERERS DRIVE GLOBAL ECOSYSTEM WORTH $82BN

Chinese money laundering networks processed 16 billion dollars in 2023, driving the global illicit ecosystem to 82 billion dollars.

ETHEREUM RESEARCH PROPOSE FOCIL AS CENSORSHIP-RESISTANCE FOR HEGOTA

Ethereum researchers introduced FOCIL to bolster censorship resistance within the 2026 Hegota upgrade.

$11 TRILLION ON THE RAILS: ETHEREUM'S TOKENIZATION MOMENT

ARK Invest projects tokenized real-world assets growing from $19B today to $11T by 2030.

Start shipping with v0

creating on v0

EARLY LOOK SHOWS APPLE'S LIQUID GLASS-STYLE BLUR EFFECTS COMING TO ANDROID 17

Google is testing blur effects in Android 17 that resemble Apple's Liquid Glass design, with translucent backgrounds appearing in system interfaces like the volume panel. The frosted visuals are tied to new UI flags and create a layered appearance in which content appears to float above blurred backgrounds. The feature remains under development and may change before Android 17's final release, alongside other updates, including private contact sharing and motion sickness reduction tools.

GOOGLE PHOTOS BRINGS PROMPT EDITING TO INDIA, AUSTRALIA, AND JAPAN

Google is expanding prompt-based AI editing in Google Photos to India, Australia, and Japan, letting users make natural-language edits on everyday photos—on-device, across many Android phones, and in multiple local languages—while adding C2PA credentials for transparent AI attribution. The move matters because it brings fast, private, conversational editing to huge mobile audiences, positioning Google against Adobe, Apple, and Samsung with a simpler, more accessible workflow despite some early limits with complex scenes.

LEAKER MAKES UNLIKELY CLAIM ABOUT IMPROVED IPHONE 18 TELEPHOTO PERFORMANCE

A leaker claims Apple is testing variable-aperture lenses and even teleconverters for the iPhone 18's telephoto camera, but the report is vague and widely seen as unlikely. While variable apertures could offer slightly better depth-of-field control on Pro models, the teleconverter claim is probably a mistranslation or misunderstanding rather than a real feature.

HOW TOP COMPANIES ARE USING AI IN THEIR DESIGN WORKFLOWS

Leading tech companies are implementing structured AI workflows in design: Atlassian uses pre-built templates and instruction files to control AI prototyping, while Meta designers automate execution tasks while maintaining manual processes for research and strategy. Teams at Tesco are building custom Figma plugins that connect to live website data, and Faire deployed an internal AI chatbot called Fairey to analyze user queries and synthesize research interviews efficiently.

ON CODING AGENTS AND THE FUTURE OF DESIGN

Coding agents like Claude Code represent a new phase of responsive design by stripping applications down to their essential primitives. This shift forces organizations to clarify what their products actually do, much like mobile-first design revealed that simpler interfaces often better serve user needs. This agent-driven future elevates design to pure strategy, where competitive advantage comes from honestly expressing capabilities rather than cluttering interfaces with organizational complexity.

IS THE DOOMSDAY CLOCK THE MOST IMPORTANT DESIGN OF OUR TIME?

The Doomsday Clock, now set at 85 seconds to midnight, endures as one of the most powerful pieces of graphic design ever made, distilling complex existential threats—nuclear war, climate change, and AI risk—into a simple, universally understood metaphor. Though subjective and imprecise, its strength lies in provoking awareness and urgency, proving that design's impact isn't accuracy but its ability to focus global attention and spur debate.

RESEARCH TOOL MADE EASY

PingPolls makes surveys simple, beautiful, and powerful with AI that helps you design, distribute, and analyze responses.

DESCRIBE AND BUILD A DEVICE

Generates a parts list, printable enclosures, firmware, and assembly steps from a text description.

GOOD WEBSITES

A curated collection of inspirational good websites.

I LEARNED THE FIRST RULE OF ARIA THE HARD WAY

A developer added role="link" to a button element, intending to style it differently, which broke keyboard activation and caused screen readers to announce conflicting roles, even though it passed automated accessibility tests. Removing the ARIA attribute and using a CSS class instead restored proper functionality, demonstrating that semantic HTML elements already provide necessary accessibility features. ARIA should only supplement native elements to convey dynamic states, such as aria-expanded, rather than override their inherent roles and behaviors.

STILLS 2026 TRENDS REPORT POINTS TO THE RISE OF HUMAN-CENTRED DESIGN

In a sea of AI-polished sameness, Stills' 2026 Photography in Design Trend Report argues that work cuts through by embracing play, imperfection, texture, and risk—making humanity visible again through colour, detail, and intentional weirdness. Across trends from scrapbook aesthetics and cyber goth to maximalism, grunge, surrealism, and sensory storytelling, the message is clear: audiences crave bold, tactile, emotionally resonant work that feels made by people, not machines.

NEW LAUNDRY BRAND EAT DIRT PRIMED TO SHAKE UP THE ECO-FRIENDLY MARKET

Former ad execs Jordan Woolley and Catherine Barr have launched Eat Dirt, a plant-based, biodegradable laundry detergent positioned as an eco-friendly yet design-forward alternative, packaged in bold, illustrated tins meant to be seen rather than hidden. Aiming at environmentally and design-conscious consumers, the brand uses irreverent naming, playful illustration, and premium pricing to shake up a conservative category and reach buyers beyond traditional supermarket channels.

DO RECRUITERS REALLY CARE ABOUT THE RESUME DESIGN?

Resume design significantly impacts recruiter perception, as messy layouts signal poor attention to detail and prompt immediate rejection during the critical 6-second initial scan window.

TASTE WILL BE THE NEW CREATIVE SUPERPOWER IN 2026

Taste and human discernment will become the defining creative differentiators in 2026 as AI-generated content floods digital spaces with low-quality "slop."