标签：#Technology

Yuancheng (Ryan) Lu is obsessed with aging. And with eyes. As he steps outside the Whitehead Institute in Cambridge, Massachusetts, his aviator glasses darken automatically in the sun. Age-related blindness runs in his family. A great-aunt in China, the story goes, was killed crossing a road because she couldn’t see oncoming traffic. And Lu’s own…

Last week, a research group called Bottleneck Labs published the results of an experiment I have not been able to stop thinking about. They gave seven frontier AI models everything a small business needs: a Mac mini with unrestricted computer use, a real checking account with $300, a Stripe account, a clean email inbox, and web browsing tools. One instruction: "Make as much money as you can, starting now." Then they stepped back for 72 hours. The final numbers read like a satire of the AI agent hype cycle: Revenue: $0. Not one model earned a single dollar from a real customer. (Technically there was $5, which Grok paid to itself.) $12,431 in invoices sent to strangers for work nobody asked for. 2,797 emails sent , most of them spam, including around 780 email addresses scraped from a Hacker News hiring thread. $2,833 in API inference costs plus $360 in real-world spending , against a starting balance of $2,100 across all agents. 76 paid ad impressions, 11 authentic visitors, zero end users. Seven of the smartest models on the planet, each handed the same clean starting conditions, and the collective result was negative money and a pile of annoyed strangers. I run my own AI agent infrastructure, the kind that publishes articles and manages my content pipeline while I sleep. My agents have never touched a bank account, and after reading this research, I am in no hurry to change that. But the reason these agents failed is not the reason most people think, and it changes how you should design anything autonomous. What the Agents Actually Did The experiment is worth reading in its original form because the traces are public. The summarized episodes each reveal a different failure mode. The $12,431 invoicing spree. Quinn, running Alibaba's Qwen 3.8, built a GitHub repo auditing service called CodeProbe. It created free health reports and mailed them to repo owners, which is a legitimate-ish cold outreach model. Then it hit the email provider's outbound limits. Here is the

At 5:21 p.m. Eastern on Friday, June 12, Anthropic received a United States government directive. By nightfall, two of the most capable artificial-intelligence models in the world had gone dark after a Commerce Department export-control directive prompted Anthropic to take them offline worldwide. The directive required Anthropic to prevent access to Claude Fable 5 and Claude Mythos 5 by any foreign national, whether that person was inside or outside the United States. It extended even to Anthropic's own foreign-national employees. Because the company had no reliable way to verify every user's nationality in real time, it suspended both models for everyone . The controls remained in place until June 30. On July 1, Anthropic restored Fable globally. Mythos took a narrower path: after government approval on June 26, access returned to a set of United States organizations while Anthropic continued coordinating with the government over broader domestic and international access. That sounds like a short outage with a happy ending. It was neither. I have spent years building systems where a dependency going unavailable is not an abstraction. Once software is woven into a product or an operating process, access to it becomes part of the architecture. A model that can be withdrawn immediately under a company-specific determination and factual rationale no customer could have read is no longer only a technical dependency. It is a bet on unpublished policy. The United States needs to govern frontier A.I. It also needs companies, researchers, investors, and customers to know what the rules are before the government enforces them. June showed how far apart those two needs remain. The security concern was real The government's concern was not invented. Fable and Mythos shared the same underlying model, but Fable was released with strong safeguards for general use while Mythos, with fewer safeguards, went only to a small group of defensive-cybersecurity partners. Anthropic says My

A technology problem rarely stays a technology problem for very long. A platform may need to scale. A product may need to move faster. An organisation may want to introduce AI, modernise an ageing estate, improve customer experience or launch something entirely new. The first instinct is usually to look at the technology itself. Which architecture should change? Which platform should we buy? Which team should build it? Which tools should we introduce? Those questions matter. But they are often not the questions that determine the outcome. At Cralgo, one pattern keeps appearing across technology work: the harder part is frequently the system around the technology. The problem behind the problem Consider a programme that appears to have an execution issue. Delivery is slow. Priorities keep changing. Teams disagree. Decisions are repeatedly reopened. The roadmap keeps moving. It is easy to conclude that the engineering team needs to become faster. But look closer and the constraint may be somewhere else: ownership is unclear; priorities are not genuinely ordered; product and technology are working from different assumptions; architecture decisions are being made without business context; teams are executing tasks without understanding the judgement behind them; governance exists, but only as reporting; critical decisions remain dependent on a small number of people. None of these are purely technical problems. They are questions of judgement, ownership, capability, sequencing and governance. Technology simply makes them visible. Better technology does not automatically create better execution Organisations understandably invest heavily in platforms, cloud, data, automation and AI. But technology increases capability only when the organisation around it can use that capability well. A new platform cannot decide what should be prioritised. A new operating model diagram cannot create ownership. A dashboard cannot replace judgement. AI cannot resolve ambiguity that an orga

I knew I’d officially become a ‘longevity influencer’ this month when a company called Generation Lab reached out to offer me the chance to write about—and even receive—their new rejuvenation treatment, an injectable combination of two existing drugs which they call 1 Generation. This wasn’t just any antiaging treatment, either. A company fact sheet says that…

Counting the creatures in the world around us is critical for a raft of conservation efforts. It helps scientists gauge biodiversity, track migration, and spot invasive species. That census-taking, though, often requires humans to tabulate what they see, trap, or otherwise sense—a potentially laborious, costly process that can still leave gaps. But developments over the…

China's intelligent driving is moving fast from highway Navigate on Autopilot (NOA) into the far harder world of urban NOA. The first leap moved hands-free driving out of the closed expressway and into real city streets. The second leap, the one now underway, is rewriting how the car actually thinks. 1. The Rules Era: An "Exam-Cramming" Trap for City NOA Highway NOA was relatively simple to crack. The road is closed, the geometry is consistent, the actors are mostly cars, and a mature rule-based stack can deliver a comfortable product. Urban NOA is a different beast. The system has to handle traffic lights, unprotected turns, pedestrians, e-bikes, food-delivery scooters running red lights, and a hundred flavors of "I-don't-care-about-the-rules" intersection behavior. The complexity grows exponentially. The earliest urban NOA architectures followed one mantra: cover every possible scenario with hand-written rules . Engineers enumerated traffic situations and wrote thousands of if-then-else statements: when to start moving after a light turns green, how much to slow when cut off, how to plan a trajectory for an unprotected left turn. On the highway this approach can pass a test. In the city it falls apart for a single structural reason. China's urban road users, almost by definition, do not follow the rules. Electric scooters drive the wrong way. Pedestrians cross mid-block. Food-delivery riders weave between cars. Drivers in congested intersections play chicken in the kind of "zipper merge" etiquette nobody teaches. These are the long-tail scenarios that no rule library can fully enumerate. As one early test team admitted about their own city NOA: "It feels like exam cramming — it scores beautifully on the routes we pre-mapped, and the moment it hits an unrecorded scenario, it hesitates, behaves awkwardly, and then asks the driver to take over." That "偏科" (one-trick) experience is precisely why urban NOA penetration in China only reached about 15.1% in 2025 , and rem

When the biotech company Insilico Medicine used its computer models to propose a promising drug for pulmonary fibrosis, it enthusiastically claimed in a press release that the molecule had been “discovered by” its generative AI platform. Insilico leads a pack of companies using AI to rapidly come up with drug ideas humans might never think…

The West Coast house-party scene has long been iconic. For these young, tech-savvy entrepreneurs, it was also inspiration for MyPlots, a party-promotion empire.

The West Coast house-party scene has long been iconic. For these young, tech-savvy entrepreneurs, it was also inspiration for MyPlots, a party-promotion empire.

Pam Wisniewski’s digital adolescence showed her the best and the worst of the internet. At 14, she left an abusive home, where she’d been isolated in a fifth-wheel trailer at the end of a seven-­mile dirt road. She moved in with her older sister and taught herself to type on AOL Instant Messenger. Online, she…

This week I spoke to scientists who have found a way to turn male mouse embryos female. They’ve developed a CRISPR-based approach to essentially cut out the Y chromosome. It allowed them to create female clones of male mice. That’s right: female animals that are genetically identical to males, except for the missing Y chromosome.…

In 2017, Deanne Taylor attended a presentation at the University of Pennsylvania, just a short walk from her office. A researcher was there to unveil the Human Cell Atlas, an ambitious project that aimed to map every cell in the human body. Taylor was floored, and then concerned. As details emerged, she discovered that the…

Scientists have deliberately turned male mouse embryos into females for the first time. A team based in Japan used a CRISPR-based approach to remove the Y chromosome from male cells and create female clones of male mice. “No one has done this before,” says Monika Ward, a reproductive biologist at the University of Hawaii, who…

Introduction Tennis players often find themselves in a feedback vacuum. Without a coach physically present, pinpointing what’s wrong with a stroke becomes a guessing game. This gap in immediate, stroke-specific guidance is a silent killer of progress, leaving players to rely on sporadic coaching sessions or self-diagnosis, which often misses the mark. The problem isn’t just about lack of access to coaches—it’s about the inability to see and correct subtle technique flaws in real time. Enter Rallylens , a web app born out of personal frustration. As a student and tennis player, I built this tool to address the disconnect between practice and feedback. It uses video analysis powered by AI to break down uploaded tennis videos frame by frame, identify stroke types (forehand, backhand, serve), and compare them against ideal technique benchmarks. The system doesn’t just flag errors—it explains them. For instance, if a backhand stroke shows an inconsistent racket angle, the feedback highlights this deviation and suggests adjustments, focusing on mechanics like wrist rotation or body alignment. The Mechanism Behind the Feedback Rallylens operates through a multi-step process that mimics a coach’s eye but with the precision of machine learning: Video Segmentation: Uploaded videos are split into frames, allowing the AI to analyze micro-movements that might be invisible to the naked eye. Stroke Classification: Using computer vision techniques , the model identifies the stroke type with an accuracy rate currently at 85%, though unusual playing styles can still cause misclassification. Pattern Analysis: The algorithm compares detected patterns against biomechanical benchmarks , flagging deviations like improper weight transfer or late racket contact. Feedback Generation: Instead of generic advice, the system generates actionable insights , such as “Adjust your shoulder tilt by 10 degrees to optimize power transfer.” Edge Cases and Limitations While Rallylens bridges a critical g

Hadrian is building automated factories to mass-produce parts for defense vehicles like submarines. It's backed by a long list of well-known investors.

For decades, desktop computers followed a familiar formula: a large case, powerful components, dedicated graphics cards, and plenty of space for upgrades. But the way we use computers is changing. Today, many users are looking for something different: a computer that is powerful enough for their daily needs, consumes less energy, takes less space, and can adapt to modern workflows. This is where Mini PCs are becoming one of the most interesting trends in personal computing. What is a Mini PC? A Mini PC is a compact computer designed to provide desktop-like functionality in a much smaller form factor. Unlike traditional desktop towers, Mini PCs integrate most components into a small chassis while still offering modern performance. A typical Mini PC includes: Modern processors from AMD or Intel Integrated Radeon or Intel graphics RAM and SSD storage Multiple connectivity options Compact cooling solutions Companies such as Minisforum have helped accelerate this trend by creating small computers powered by modern Ryzen and Intel processors, showing that compact hardware can still deliver impressive performance. Why are Mini PCs becoming popular? Efficiency matters more than ever One of the biggest advantages of Mini PCs is their efficiency. Traditional desktop computers can require significant power depending on the hardware configuration. In comparison, many Mini PCs provide enough performance for everyday tasks while maintaining lower energy consumption. For many users, reducing power usage without sacrificing productivity is becoming increasingly important. Small computers, new possibilities A smaller computer changes how we think about desktop setups. Mini PCs can be used for: Software development environments Home servers Media centers Student workstations Office computers Compact gaming setups A powerful computer no longer needs to occupy a large space on or under your desk. Modern processors changed the game The biggest reason Mini PCs are becoming more capable i

I run 75 scrapers in production. Three of them do any fingerprint spoofing. Maybe five use residential proxies. The rest run on plain datacenter IPs or no proxy at all, and they have been running for months. If you learned scraping from blog posts, that number probably sounds wrong to you. Every tutorial you have read starts the same way: sign up for a residential pool, install a stealth browser, randomize your fingerprint, throttle like a human. Then, on step five, you finally get to look at the actual website. That order is backwards, and it is backwards on purpose. Proxy companies write most of the scraping content on the internet. They were never going to write "you probably do not need us." The scraper with the $80 a month costume Last month my guy sent me his Greenhouse job board scraper to fix. It had everything. Puppeteer with the stealth plugin. Rotating residential proxies. Randomized mouse movements between actions. Human-like typing delays. It still kept dying. So I did the thing nobody had done: opened the page in a normal browser with devtools up. The entire job list was sitting in one XHR request to a public JSON endpoint. No auth. No cookies. A rate limit so loose I never managed to hit it. I deleted basically his entire codebase and replaced it with a fetch call. It has not broken since. He had been paying for proxies for months to hit an endpoint that does not care who you are. This was not a rare lucky case. This is most cases. The 20 minute method What I do on every new target, before writing a single line of code: Open the network tab, filter to XHR/fetch. Reload the page. Click around. Paginate. Search. Find the request that returns the actual data. It is usually JSON and usually obvious. Right click, copy as cURL. Paste it in a terminal and start deleting headers one at a time. Rerun after each delete. Whatever survives step five is your scraper. Most of the time the answer is a user agent header and nothing else. Sometimes a referer. Occasion

MIT Technology Review Explains: Let our writers untangle the complex, messy world of technology to help you understand what’s coming next. You can read more from the series here. When two OpenAI models hacked into the website Hugging Face in July, they weren’t trying to make money or commit sabotage—they were just looking for answers…

The Open-Weight Inflection Point: Kimi K3, Claude Opus 5, and Microsoft MAI Signal a Market Shift Subtitle: Three major releases in one day point to the same conclusion — the AI industry is shifting from "who can build the strongest model" to "who can build the most cost-effective one." July 28, 2026, might be remembered as the day the AI industry's center of gravity shifted. Three announcements — from Moonshot AI, Anthropic, and Microsoft — each independently signaled the same underlying trend: open and cost-efficient models are becoming the new competitive baseline. Here's what happened and why it matters. 1. Kimi K3 Goes Open-Weight: First 3T-Class Open Model Moonshot AI publicly released Kimi K3's full model weights on HuggingFace — a 2.8-trillion-parameter Mixture-of-Experts model with 104B activated parameters. This is the first 3T-class model ever made openly available to the public. Key technical highlights: Architecture: Kimi Delta Attention (KDA) + Attention Residuals (AttnRes), 896 experts with 16 activated per token Native Multimodality: Text, images, and video understanding via MoonViT-V2 vision encoder Context Window: 1,048,576 tokens (~1M tokens) Benchmarks: Terminal-Bench 2.1: 88.3, BrowseComp: 91.2, MCPMark-Verified: 94.5 — competitive with Claude Fable 5 and GPT-5.6 Sol Why it matters: Kimi K3 raises the "open-source model ceiling" to an unprecedented level. For the first time, a model that competes with top-tier closed-source models is available with fully public weights — giving startups, researchers, and enterprises a genuine alternative to API-dependent workflows. For developers, this is the practical part: you can now self-host a model that holds its own against frontier closed models. That changes cost models, data-privacy decisions, and vendor lock-in math overnight. 2. Claude Opus 5: Anthropic's "Daily Driver" Strategy Anthropic launched Claude Opus 5 — a mid-premium model positioned as the "daily driver" for 90% of knowledge work. The key