📌 TOPINDIATOURS Hot ai: Nous Research's NousCoder-14B is an open-source codin

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

📌 TOPINDIATOURS Update ai: Tech Billionaire Says It’s Time for the Government to S

From the Alien and Sedition Acts to various Red Scares, American history is littered with panicked attempts to curtail the First Amendment to defend the status quo. Israeli multi-billionaire Shlomo Kramer is the latest to make the case — arguing that the country must control its citizens’ speech in order to survive, an astonishingly bold assertion from someone who isn’t even a US citizen.

In a recent interview with CNBC, Kramer argued that free speech protections are leading the US down the wrong path.

“You’re seeing the polarization in countries that allow for the First Amendment and protect it, which is great — and I know it’s difficult to hear, but it’s time to limit the First Amendment in order to protect it,” he told the CNBC hosts. “And quickly before it’s too late.”

Kramer, whose perspective is forged by his time in the Israel Defense Forces’ counterintelligence group Israeli Unit 8200 — which has been called a “secret startup machine” for its track record of producing tech founders — has some hefty credentials in the business world. He cofounded his first major security company, Check Point Software Technologies, along with alumni from Unit 8200, and later built and sold web security firm Imperva for $3.6 billion. Now CEO of cloud security company Cato Networks, a firm partnered with major US corporations like Amazon Web Services and RingCentral, Kramer is wading into American politics with some radical proposals.

Asked to clarify exactly what he meant about limiting the First Amendment, Kramer told CNBC that the government must take control of social media to avoid negative polarization — in other words, online debate.

“I mean that we need to control the platforms, all the social platforms,” he explained. “We need to stack, rank the authenticity of every person that expresses themselves online and take control over what they are saying, based on that ranking.”

“The government?” CNBC host Sara Eisen asked.

“The government should, yeah,” Kramer confirmed. “They should do that. And we need to educate people against lies.”

It’s hard to separate the billionaire’s outlook from the secular nationalism which he advocates for in his own country — a perspective that prioritizes state security and market-driven solutions over concerns like civil liberties.

For example, 2024 polling by the Pew Research Center found that at least half of all Israeli adults surveyed were in favor of social media censorship, particularly when it comes to content about the country’s controversial war in Gaza. More recent analysis by the Committee to Protect Journalists found that Israeli journalists face escalating censorship, harassment, and pressure to self-censor as a decade-long propaganda campaign by Netanyahu’s government reaches a fever pitch.

There’s another messy implication: such a vast censorship campaign would require an extensive cybersecurity apparatus — which the tech billionaire just so happens to be in the business of supplying.

“You need to put adjustments that are perhaps not popular, but necessary,” Kramer said. “This is an urgent need, by the government, to do that. And until then, enterprises are buying — by themselves — more and more cybersecurity solutions… but they are looking for ways to drive more efficient consumption.”

“And I assume you have some of those ways, obviously,” CNBC host David Faber replied.

“So, that drives the next generation of companies, such as Wiz, Crowdstrike, the Cato Networks,” Kramer continued, referencing his own security firm. “Networks that are platforms, and are able to deliver this extended need for security in affordable ways for enterprises.”

More on billionaires: Tech Billionaires Are Starting Private Cities to Escape the United States

The post Tech Billionaire Says It’s Time for the Government to Suspend Freedom of Speech appeared first on Futurism.

🔗 Sumber: futurism.com