标签：#c

Valve has been incredibly friendly to customers who need repairs - which is why it was so surprising to hear that Valve was already discontinuing the battery for the Steam Deck LCD handheld. It would have meant you could no longer just buy a battery from iFixit and install it yourself. Why would Valve stop […]

While an impressive sale, the pen and switch did not break records.

If you have ever built an autonomous mobile robot, you have likely run into the dreaded "Shuffled USB Port" problem. You boot up your robot, fire up your ROS 2 launch files, and... crash. Your LiDAR driver is trying to parse data from your IMU, and your IMU node is screaming about invalid serial frames. Because Linux assigns virtual serial paths like /dev/ttyUSB0 and /dev/ttyUSB1 based purely on which device initialized milliseconds faster, relying on default OS paths is a recipe for system instability. When you are scaling up to dozens of Jetson Orin nodes —each equipped with an RPLIDAR C1 and a Yahboom 10-axis IMU —manually hardcoding paths or writing rigid scripts on every individual machine isn't viable. Here is how production-grade robotics fleets handle plug-and-play USB binding dynamically using configuration-driven udev rules. The Core Concept: Vendor ID vs. Physical Port vs. Serials Linux's udev (device manager) allows us to dynamically create stable symbolic links (symlinks) like /dev/rplidar and /dev/imu when hardware is plugged in. How we identify those devices determines our fleet's flexibility: USB Serials: Unique to each individual chip. Highly secure, but requires registering every single replacement sensor in your codebase. Physical USB Ports ( KERNELS ): Tied to a physical slot on the carrier board. Great if you have identical sensors, but forces technicians to plug cables into highly specific, undocumented ports. Vendor ID (VID) & Product ID (PID): Identifies the USB-to-serial converter chip on the sensor board. Because the RPLIDAR C1 uses a Silicon Labs CP210x chip ( 10c4:ea60 ) and the Yahboom IMU uses a QinHeng CH340 chip ( 1a86:7523 ), they use completely distinct silicon. This means we can map them dynamically and reliably using just their VID/PID —allowing field technicians to plug them into any USB port on the Jetson without breaking the system. Step 1: The Configuration-Driven File ( devices.conf ) Hardcoding vendor rules inside shell scri

The conveneince store chain will use StorMagic instead.

It's likely to become a direct power source for SpaceXAI's data centers.

The Elon Musk-owned xAI is suing a South Carolina man who allegedly used the company's Grok AI chatbot to generate child sexual abuse material (CSAM). In a lawsuit reported earlier by Reuters, xAI claims Terry Wayne Harwood "knowingly and intentionally used Grok to circumvent safeguards, alter nonconsensual images, and generate and distribute CSAM," breaching the […]

For five years, the answer to "how do we make the model better" was always the same: bigger model, more data, wait for the loss curve to bend. Scaling laws made AI progress feel almost boring — predictable, like compound interest. That knob still works. But it's not the interesting one anymore. The real question teams are fighting with now is: given a model you already have, how much compute should it spend per question? Not per training run. Per question. And the answer turns out to break almost every assumption your inference stack was built on. The single forward pass was always the weak link Here's the thing nobody says out loud enough: a model's first answer is often just a noisy sample. The right answer was in there somewhere — it just wasn't the loudest one. You don't fix that with more pretraining tokens. You fix it by sampling more, checking the work, or letting the model take a second pass. That's the whole idea behind test-time compute. And there are three ways to spend that extra compute, and they are not interchangeable: Let it ramble longer. Cheap. Also the reason some models talk themselves out of the right answer after 3,000 tokens of unnecessary detours. Generate N answers, vote. Parallel, simple, plugs into infra you already have. Also: returns drop off fast. Past a certain N you're burning GPU-seconds for crumbs. Search a tree of partial reasoning steps, prune the bad branches early with a verifier. The most compute-efficient option by far. Also the one that quietly breaks your serving stack. Most real systems end up doing all three — cheap by default, escalating only when a confidence check says "this one's hard." That escalation decision is its own piece of architecture. Most teams bolt it on as an afterthought. It shouldn't be one. Your serving stack was not built for this Continuous batching, paged KV-cache, speculative decoding — all of that infra assumes requests are independent, similar in length, and finish in one pass. Test-time compute v

The Federal Communications Commission will vote next month on whether a single company can own broadcast stations that reach more than 39 percent of US TV households. In a Breitbart op-ed on Wednesday, Republican Chair Brendan Carr announced an August 6th vote to end the national ownership cap rule, which was intended to prevent one […]

Disinformation researchers praise ruling blocking Trump visa denials and deportations.

A 90% discount doesn't come from nowhere. Someone pays for it, and usually that someone is you: with your data, with model quality, or with somebody's stolen account. An entire industry has grown up around Anthropic and OpenAI over there. They call them "transfer stations": proxies that resell tokens from Western models. This stopped being a garage workaround for rate limits a long time ago. It's a market with its own economics, and it runs on a few things. Accounts. These services mass-produce free and Max subscriptions through stand-in registrants and other people's KYC, then wrap them in a pseudo-API. Model swapping. You pay for the flagship, but your simple request gets quietly routed to something cheaper. Who actually generated the answer? You can't check. Data. The proxy operator sees all of it: prompts, responses, code, internal context. Even if they never sell the logs, you don't know where those logs sit, who reads them, or whether they ended up in somebody else's training run. Take Yunwu. By their own price list, access to Anthropic models runs up to 93% below official. The price alone proves nothing. But the market logic is simple: restrictions don't kill demand, they make it more expensive, and they feed the workaround industry. In June 2026 Anthropic accused entities tied to Alibaba/Qwen of a campaign to siphon Claude: nearly 25,000 fake accounts and 28.8 million requests in six weeks, to distill their own models. DeepSeek, Moonshot and MiniMax have caught similar accusations before. Same grey access, just wholesale. But the problem is bigger than China. Even with fully legitimate access, most teams can't answer three questions about their own production: which model actually served the request, what it cost, and whether quality dropped. The grey market just takes that blindness to its logical end. I think the priorities here are upside down. For a serious product, production visibility and control over your data matter more than a few percent saved on