AI Weekly Malaysia

Meta released Muse Glimmer, a 30-billion parameter open-weight model under Apache 2.0 designed to run AI agents locally on a single consumer GPU (Mac or PC). It supports text and images, was trained across 100+ languages, and handles multi-step agentic tasks like tool calling, code writing/debugging, and file/screenshot manipulation, working offline as an 'always-on' personal agent.

Why: A 30B parameter agentic model that runs on a single consumer GPU under Apache 2.0 is directly downloadable and deployable today — builders can prototype local AI agents without cloud API costs or data leaving the device. For Malaysian developers and startups, this matters because local execution sidesteps cloud latency and data residency concerns, and the 100+ language training may include Malay or other regional languages worth testing. Evaluate whether Glimmer's agentic capabilities (tool calling, code debugging, file handling) are good enough to replace or complement your current cloud-based agent stack.

Needle 2 is a 45M-parameter agentic LLM compressed to a 14MB binary that runs in 28MB of RAM, designed for tool calling, device control, and structured extraction on sub-$200 hardware. It hits 500 tokens/sec decode on a Raspberry Pi 5, 300–700 on budget Samsung A-Series phones, and even runs on ESP32-S3 microcontrollers. It trades benchmark wins with models 5×–70× larger (FunctionGemma 270M, LFM2.5 230M, Apple FM) on mobile device-use tasks, and is released under Apache 2.0 with weights on Hugging Face.

Why: If you are building IoT, smart home, wearable, or robotics products targeting the Malaysian or broader SEA market where most phones ship under $200, Needle 2 lets you run on-device function calling without a GPU, NPU, or cloud dependency. The 28MB RAM footprint means you can prototype agent-based device control on hardware you already have—Raspberry Pi, ESP32-S3, or budget Android phones—today, with the repo and sandbox available to test immediately.

A new paper from Multiverse Computing introduces two systems-level optimizations for LLM knowledge distillation: caching the teacher's top-K logits offline so the teacher model never needs to co-reside in VRAM with the student, and a fused chunked KL-divergence loss that avoids materializing the full vocabulary×sequence-length probability matrix. Together these cut VRAM usage far below default PyTorch or NVIDIA Megatron-Bridge implementations, making long-context distillation feasible on a single GPU instead of requiring hundreds.

Why: If you're distilling large open-source models (e.g., gpt-oss-120b with its 201,088-token vocabulary) into smaller deployable students, this approach lets you skip keeping the teacher loaded during training—potentially dropping your GPU footprint from a cluster to a single card. Evaluate the offline top-K logits caching and fused KL loss before your next distillation run, especially if you've been blocked by VRAM costs.

Nathan Lambert reflects on writing an AI textbook and argues that LLMs remain stagnant at long-form non-fiction writing, increasing entropy rather than compressing knowledge into insight. He contends that if models can't organize and present established science, they're not ready to autonomously solve open-ended scientific problems, and that progress will look more like low-hanging fruit and cross-field connections than revolutionary breakthroughs.

Why: If you're building AI agents for research, technical writing, or autonomous knowledge work, this argues against assuming models will soon self-organize complex information into coherent long-form output. Plan for human-in-the-loop structuring and editing rather than end-to-end autonomous generation for anything requiring sustained argument or knowledge compression.

LiquidAI released LFM2.5-VL-3B, a 3.1B parameter vision-language model designed for on-device/edge use, pairing a SigLIP2 400M vision encoder with their LFM2.5-2.6B text backbone. It was pre-trained on ~34T tokens with 4x more vision data than prior versions, supports 128K vocabulary for non-Latin scripts, and adds screen/UI understanding, object grounding, multi-image input, and function calling. Benchmarks show it leading its size class on real-world image tasks (RealWorldQA 73.1, MMStar 63.3) against comparably-sized models from Qwen, InternVL, and Gemma.

Why: If you are building on-device apps that need document/screen understanding or vision-grounded function calling without cloud API latency or cost, this is a concrete 3B model worth benchmarking against Qwen3.5-2B or InternVL 3.5 2B for your use case. The function-calling capability in vision-text contexts is the differentiator to test, since most small VLMs struggle there.

Malaysians with passports valid for more than 10 years—caused by JIM rolling over remaining months when renewing early—were barred from international flights because airline booking systems reject the extended validity dates. JIM has capped all new passports at exactly 10 years and is offering free replacements for affected holders, plus priority service at immigration offices.

Why: If you hold a Malaysian passport with validity exceeding 10 years, you should apply for a free replacement before your next international trip to avoid being denied boarding. For builders, this is a concrete example of a data validation edge case in airline systems forcing a government policy change—a useful case study in how rigid field constraints in booking systems can create real-world friction with legitimate documents.

A routine cyber vulnerability assessment found cameras aboard Royal Navy Kraken Unmanned Surface Vessels transmitting heartbeat signals to an IP address in China. The UK MoD confirmed no classified data was compromised, and the cameras were sourced by Kraken from a third-party supplier. The incident underscores that even benign telemetry transmissions from unvetted supply chain components can surface as serious security incidents.

Why: If you ship IoT or edge devices, or build systems that incorporate third-party hardware, this is a concrete reminder to network-monitor your own components for unexpected outbound traffic before a customer or auditor finds it. The failure here wasn't malware—it was an unexamined camera phoning home. Builders integrating hardware from opaque supply chains should demand component-level network behavior documentation and run passive traffic capture during staging, not trust supplier assurances.

Ukraine claims to have found an Nvidia Jetson edge-AI chip inside a Russian S-71 'Monochrome' cruise missile, suggesting the weapon may use AI for targeting or navigation. The Jetson line is designed for embedded and edge inference workloads, making its presence in a munition notable as an indicator of AI-enabled autonomous weapons systems.

Why: If confirmed, this is a concrete data point that edge-AI inference hardware (the same Jetson modules developers use for robotics and vision projects) is being deployed in autonomous weapons — relevant to anyone building with Jetson who should be aware of export-control and dual-use scrutiny that may tighten around these chips.