MAROKO133 Hot ai: Red Hot Christmas Toy Crashes So Badly That Kids Can’t Actually Use It E

📌 MAROKO133 Breaking ai: Red Hot Christmas Toy Crashes So Badly That Kids Can’t Ac

Parents were left scrambling on Christmas day after the hot toy of the season appeared to crash and burn, in a perfect cautionary tale about the era of connected gadgets that can easily brick if their delicate infrastructure is put under strain.

The toy, called the Tin Can, is basically a stripped down landline phone that places calls over WiFi. The devices are styled after colorful tin cans — like an old school tin can telephone, an archaic toy so thoroughly ancient that kids today presumably have no idea it ever existed — with twirly cords just like the landlines of yore.

The pitch is that the Tin Can allows parents to set up a sort of closed network of other Tin Can users, making it a nice compromise for kids who are too young for a cell phone, but who still want to communicate with their friends and family members. It’s free to call other Tin Can users, or parents can pay a $10 monthly subscription to make outside calls.

Yet when kids rushed to phone their friends over the holidays to schedule playdates and sledding runs, they found the string had been cut.

“Call volume on Christmas Day increased more than 100x from the start of the month, which impacted people’s abilities to set up their devices or make calls,” Tin Can’s founder Chet Kittleson told Business Insider in an interview. “Despite spending months and months preparing for it, we didn’t get it all right.”

“It worked great, the kids were using it,” Maria Pahuja, a parent from Virginia, told the publication. “Then, Christmas morning, it stopped working. Sometimes you’d pick it up, and there would be a dial tone, you would call, and nothing would happen. Once in a while, a call would go through, and then two minutes later another call wouldn’t.”

In other words, kids probably would have had a better chance of connecting with their friends with an actual tin can telephone, operated via vibrationson a string, on Christmas morn’.

By now, things seem to be back up and running, though there are still some issues with call quality popping up, BI reports. A January 8th status update on the company’s website — and, tellingly, a banner on the very top of its home page — advises parents that “we’re continuing to see very positive signs of recovery, but we’re still working through some lingering issues.”

The company began in Seattle “with a handful of families and has since grown to serve families across more than 30 states,” its site says. BI reports the company has successfully raised $15.5 million since its launch in Autumn of 2024 — resources which apparently weren’t enough to prevent the toys from turning into a corded lump of coal at the exact moment they were supposed to make a great first impression.

More on gadgets: New AI Device Pours Alcohol Directly Into the Void Where Your Soul Should Be

The post Red Hot Christmas Toy Crashes So Badly That Kids Can’t Actually Use It appeared first on Futurism.

🔗 Sumber: futurism.com

📌 MAROKO133 Eksklusif ai: Nous Research's NousCoder-14B is an open-source cod

Nous Research, the open-source artificial intelligence startup backed by crypto venture firm Paradigm, released a new competitive programming model on Monday that it says matches or exceeds several larger proprietary systems — trained in just four days using 48 of Nvidia's latest B200 graphics processors.

The model, called NousCoder-14B, is another entry in a crowded field of AI coding assistants, but arrives at a particularly charged moment: Claude Code, the agentic programming tool from rival Anthropic, has dominated social media discussion since New Year's Day, with developers posting breathless testimonials about its capabilities. The simultaneous developments underscore how quickly AI-assisted software development is evolving — and how fiercely companies large and small are competing to capture what many believe will become a foundational technology for how software gets written.

type: embedded-entry-inline id: 74cSyrq6OUrp9SEQ5zOUSl

NousCoder-14B achieves a 67.87 percent accuracy rate on LiveCodeBench v6, a standardized evaluation that tests models on competitive programming problems published between August 2024 and May 2025. That figure represents a 7.08 percentage point improvement over the base model it was trained from, Alibaba's Qwen3-14B, according to Nous Research's technical report published alongside the release.

"I gave Claude Code a description of the problem, it generated what we built last year in an hour," wrote Jaana Dogan, a principal engineer at Google responsible for the Gemini API, in a viral post on X last week that captured the prevailing mood around AI coding tools. Dogan was describing a distributed agent orchestration system her team had spent a year developing — a system Claude Code approximated from a three-paragraph prompt.

The juxtaposition is instructive: while Anthropic's Claude Code has captured imaginations with demonstrations of end-to-end software development, Nous Research is betting that open-source alternatives trained on verifiable problems can close the gap — and that transparency in how these models are built matters as much as raw capability.

How Nous Research built an AI coding model that anyone can replicate

What distinguishes the NousCoder-14B release from many competitor announcements is its radical openness. Nous Research published not just the model weights but the complete reinforcement learning environment, benchmark suite, and training harness — built on the company's Atropos framework — enabling any researcher with sufficient compute to reproduce or extend the work.

"Open-sourcing the Atropos stack provides the necessary infrastructure for reproducible olympiad-level reasoning research," noted one observer on X, summarizing the significance for the academic and open-source communities.

The model was trained by Joe Li, a researcher in residence at Nous Research and a former competitive programmer himself. Li's technical report reveals an unexpectedly personal dimension: he compared the model's improvement trajectory to his own journey on Codeforces, the competitive programming platform where participants earn ratings based on contest performance.

Based on rough estimates mapping LiveCodeBench scores to Codeforces ratings, Li calculated that NousCoder-14B's improvemen t— from approximately the 1600-1750 rating range to 2100-2200 — mirrors a leap that took him nearly two years of sustained practice between ages 14 and 16. The model accomplished the equivalent in four days.

"Watching that final training run unfold was quite a surreal experience," Li wrote in the technical report.

But Li was quick to note an important caveat that speaks to broader questions about AI efficiency: he solved roughly 1,000 problems during those two years, while the model required 24,000. Humans, at least for now, remain dramatically more sample-efficient learners.

Inside the reinforcement learning system that trains on 24,000 competitive programming problems

NousCoder-14B's training process offers a window into the increasingly sophisticated techniques researchers use to improve AI reasoning capabilities through reinforcement learning.

The approach relies on what researchers call "verifiable rewards" — a system where the model generates code solutions, those solutions are executed against test cases, and the model receives a simple binary signal: correct or incorrect. This feedback loop, while conceptually straightforward, requires significant infrastructure to execute at scale.

Nous Research used Modal, a cloud computing platform, to run sandboxed code execution in parallel. Each of the 24,000 training problems contains hundreds of test cases on average, and the system must verify that generated code produces correct outputs within time and memory constraints — 15 seconds and 4 gigabytes, respectively.

The training employed a technique called DAPO (Dynamic Sampling Policy Optimization), which the researchers found performed slightly better than alternatives in their experiments. A key innovation involves "dynamic sampling" — discarding training examples where the model either solves all attempts or fails all attempts, since these provide no useful gradient signal for learning.

The researchers also adopted "iterative context extension," first training the model with a 32,000-token context window before expanding to 40,000 tokens. During evaluation, extending the context further to approximately 80,000 tokens produced the best results, with accuracy reaching 67.87 percent.

Perhaps most significantly, the training pipeline overlaps inference and verification — as soon as the model generates a solution, it begins work on the next problem while the previous solution is being checked. This pipelining, combined with asynchronous training where multiple model instances work in parallel, maximizes hardware utilization on expensive GPU clusters.

The looming data shortage that could slow AI coding model progress

Buried in Li's <a href="https://nousresearch.com/nouscoder-14b-a-co…

Konten dipersingkat otomatis.

🔗 Sumber: venturebeat.com