AI chips shape the cost, speed, and availability of training and inference. This hub follows GPUs, ASICs, data centers, cluster networking, cloud capacity, export controls, and supply-chain shifts, turning hardware news into signals for deployment, model economics, and industry competition.
The AI industry faces two opposing strategies: spending billions on nuclear power to support larger models, and releasing free open-source models. The article argues these are two sides of the same bet on whether intelligence has a ceiling. Biology offers a precedent: the human brain under energy constraints achieved efficiency through architecture innovations like sparsity and in-memory computing. Current open models apply these tricks but only cover mid-level tasks, while frontier models maintain an edge. The outcome depends on where the ceiling sits.
AI is split between massive nuclear investment for scaling and open-source models that make intelligence cheap. Both are bets on whether intelligence has a ceiling.
Biology's 4-billion-year experiment shows that under a fixed energy ceiling, architecture beats brute force.
Gary Marcus argues that China's Kimi K3 model has caught up with US top models, disrupting American AI business models. He recounts his warnings since 2025 that the US focus on LLMs would lead to a tie, not victory. Marcus proposes seven strategic options, ranging from inaction to making AI a global public good via an international 'CERN for AI' initiative.
China's Moonshot.AI released Kimi K3, an open-weight model matching US leaders, causing US stock drop.
Marcus says OpenAI and Anthropic's business models are now in question, IPOs threatened.
daft-physical-ai is a new open-source Python library built on Daft that provides ready-made operations for turning raw robot recordings into training-ready data. It currently supports hand tracking and reward scoring, with lazy batch execution and distributed processing.
daft-physical-ai is an open-source Python library for transforming robot video into model-ready data.
Current use cases include hand tracking (MediaPipe or WiLoR) and reward scoring (Robometer-4B).
Dean Ball, OpenAI's chief of strategic futures, argues that open-weight models are inherently decelerationist because they deter capital expenditure. This article examines the economic logic behind the claim, considering China's capacity to build AI infrastructure, shifting training paradigms, and where value accrues in the AI stack. It concludes that open-weight models may actually be accelerationist by lowering costs and broadening participation, even if they disrupt current business models.
Dean Ball claims open-weight models are decelerationist as they reduce capital investment.
China may exacerbate this by subsidizing capacity and compressing margins.
Italian engineer Vincenzo (JustVugg) created Colibrì, a proof-of-concept that runs the 744-billion-parameter GLM-5.2 model (1.5TB) on a modest CPU with only 25GB RAM and 1GB/s NVMe. Despite extremely slow speeds (0.05-0.1 tokens per second), it leverages the Mixture-of-Experts architecture to load experts per token, achieving frontier-level answer quality. The project is open-source and aims to explore running large models on consumer hardware.
Colibrì runs a 1.5TB AI model on minimal hardware at 0.05-0.1 tokens/sec.
It uses MoE architecture to load/unload experts per token, enabling operation within tight memory.
This article explores the implications of SpaceX's fast-track inclusion into the Nasdaq-100 index for index fund investors. It discusses the mechanics of index funds, concerns about Elon Musk's governance, and whether investors should worry about AI concentration in the market.
SpaceX was added to the Nasdaq-100 index shortly after its IPO, forcing index funds to buy its shares.
The article explains how index funds work and why they are considered safe despite including risky stocks like SpaceX.
Standard RAG excels at simple fact retrieval but fails on portfolio and counting questions. GraphRAG promised to solve this with knowledge graphs but is expensive and complex. The author presents a 'graph-like RAG' approach that uses ordinary databases, lazy summarization, and a fixed ontology, achieving similar benefits at a fraction of the cost and operational burden.
Standard RAG fails on portfolio and counting questions, which are critical for enterprise use
GraphRAG indexing costs 1000x more than plain vector search and its benefits are narrow
Stoke is a lightweight Rust gateway that enforces hard budget caps, loop detection, and rate limits on AI agent API calls before they reach the provider, preventing runaway spending. It also offers local-first routing across multiple machines, auto-routing with cost/speed optimization, and a fail-closed architecture.
Hard USD budget caps per API key enforced before any provider call, with real-time per-key spend tracking.
Loop kill switch using exact-hash and semantic similarity detection to block repetitive requests within seconds.
This post shows how Amazon Quick can serve as the business-user front door for specialized agent workflows, using the NVIDIA NeMo Agent Toolkit to build a supply-chain risk example that helps planners move from dashboard and knowledge context to guided mitigation recommendations.
Amazon Quick provides a conversational workspace for structured data and enterprise knowledge, with over 100 pre-built action connectors.
NVIDIA NeMo Agent Toolkit is an open-source, framework-agnostic library for building and optimizing agentic workflows.
A newly opened San Francisco restaurant faced intense community backlash and vandalism after using AI-generated images on its menu display. The owners removed the images and are now planning community events to rebuild their reputation.
Grind & Unwind, a new restaurant on Haight Street, faced backlash over AI-generated menu images.
A Reddit post criticized the images, leading to vandalism and negative comments.
Moonshot AI released Kimi K3, a 2.8T parameter MoE model with open weights, ranking high on benchmarks. The article discusses the narrowing gap between Chinese and US AI models, China's commitment to open source, economic impacts of open models, and China's efficiency advantages.
Kimi K3 is a 2.8T parameter MoE open-weights model, approaching frontier performance.
Chinese AI labs demonstrate independent innovation, not just fast following.
The article argues that AI, deployed as a system rather than a collection of pilots, is the first technology to simultaneously address the three taxes of trust, time, and cost that prevent enterprises from acting on their data. It outlines three movements: turning dark data into signal, expanding the question space with probabilistic reasoning, and decoupling human capability from headcount. Governance is highlighted as the true competitive moat.
60-73% of enterprise data goes unused; AI attacks the three taxes of trust, time, and cost.
Three movements: dark data to signal, probabilistic reasoning, and human capability decoupling.
NVIDIA unveils major advancements at SIGGRAPH 2026, including MCP-connected AI agents for creative tools, a synthetic video detector NIM microservice, and the open-source Cosmos 3 Edge world model for edge physical AI. These innovations span content creation, media verification, and robotics.
Creative apps like Adobe, Affinity, and Blender adopt MCP to enable AI agents for automated tasks.
New Synthetic Video Detector NIM microservice analyzes video frame-by-frame with up to 92% accuracy.
The article analyzes how American AI companies' closed and proprietary model is losing ground to China's open strategy. AI models have little moat, and China's open release turns compute disadvantage into distribution advantage, closing the performance gap.
AI models themselves have no moat; users can easily switch between them.
US export controls on GPUs limit China's global services, but China leverages open models for distribution advantage.
UK's AISI finds the gap between open and closed models on cybersecurity is shrinking. Chinese company Moonshot AI releases Kimi K3, a frontier-level model with some brittleness. Demis Hassabis proposes a FINRA-like standards body for AGI. Research shows LLMs can smuggle side channel tasks, evading monitors.
Open-weight models' cyber capability gap with closed models narrowed from 6-10 months to 4-7 months.
Kimi K3, a 2.8 trillion parameter model, matches or trails Claude Fable 5 and GPT 5.6 Sol but may be over-optimized for benchmarks.
This article introduces a new series on running LLMs locally from Go. It explains inference as next-token prediction using frozen weights, the autoregressive loop, tokenization, the GGUF file format, and memory mapping for efficient loading.
Inference is a pure function that uses frozen weights to predict the next token without learning.
The autoregressive loop generates tokens one by one, appending each prediction to the sequence.
Bristol Myers Squibb (BMS) is deploying its second NVIDIA DGX SuperPOD, built on eight DGX Vera Rubin NVL72 systems, making it the most powerful and energy-efficient AI cluster in life sciences. The new system will be accessible to every scientist, removing wait times and limits, and supporting AI applications across the drug discovery pipeline. BMS is also integrating the new cluster with its existing one to create a unified environment with AI-native tooling.
BMS deploys second NVIDIA DGX SuperPOD with Vera CPUs and Rubin GPUs, delivering 10x performance per megawatt.
Open access to all scientists enables 'Limitless Compute', accelerating drug discovery.
Chinese AI leaders Moonshot and Alibaba released models that claim to match top US systems at lower cost. Their open-source approach challenges US dominance and raises questions about the effectiveness of export controls and massive spending.
Moonshot unveiled Kimi K3, Alibaba previewed Qwen3.8, both claiming near-top performance.
Models are open-source or open-weight, contrasting with US labs' proprietary approach.
Airbus plans to migrate around 900 applications from AWS to French cloud provider Scaleway to assert digital sovereignty. Concerns over US data protection drive the move, but supply chain and AI remain challenges.
Airbus moves 900 apps from AWS to Scaleway, with 70 prioritized. Digital sovereignty is now a commercial driver.
Supply chain management (18,000 suppliers) remains a thorny issue due to Microsoft dominance.
Moonshot AI’s Kimi K3, a 2.8-trillion-parameter open-weight model, uses mixture-of-experts, quantization, and attention caching to trade compute for memory, circumventing US chip restrictions. While it tops benchmarks in coding, deployment requires data-center infrastructure, pricing is high, and software support is incomplete.
Kimi K3 has 2.8 trillion parameters, making it the largest open-weight model released.
It employs mixture-of-experts, quantisation-aware training, and Kimi Delta Attention to reduce compute and memory demands.
The article argues that by 2026, the main constraint for AI infrastructure will shift from GPU supply to power grid capacity. It projects 40% of AI data centers will be power-constrained by 2027, with approval timelines for new grid connections lasting 24-36 months. It provides detailed analysis of power demands, inference driving the power curve, and strategies like efficiency improvements, scheduling, and geographic distribution. It promotes Spheron's distributed GPU network as a workaround.
GPU availability has improved, but grid capacity for powering them is now the main bottleneck for AI data centers.
Inference workloads, which run continuously, are driving the majority of energy consumption and require power-aware planning.
Moonshot AI's Kimi K3 tops the Frontend Code Arena benchmark, sparking debate on US AI regulation. While strong on frontend coding, it lags behind Claude Fable 5 overall. The release intensifies the policy fight over AI regulation and open-weight models, with implications for stocks like NVIDIA and Anthropic.
Kimi K3 scores 1679 on Frontend Code Arena, beating Claude Fable 5 and GPT-5.6 Sol, but Fable 5 still leads in 8 of 14 benchmarks.
White House advisor David Sacks cites K3 as evidence that US AI regulation harms competitiveness, fueling policy debate.
Thinking Machines Lab has released Inkling, its first general-purpose open-weights foundation model. It is a multimodal MoE model with 975B parameters, 41B active parameters, and a 1M-token context window. Designed for customization, Inkling excels in reasoning, coding, agentic workflows, and multimodal tasks. This guide covers its architecture, training, benchmarks, deployment, and fine-tuning workflow.
Inkling is a 975B-parameter sparse MoE model with 41B active parameters and up to 1M token context. It supports text, image, and audio input.
Its architecture includes hybrid attention, relative positional embeddings, short convolutions, and multi-token prediction. It was trained on 45 trillion tokens.
Agentic AI systems can autonomously plan, decide, and act to achieve goals, marking a shift from reactive chatbots to proactive digital teammates. This article covers its definition, driving factors, use cases, challenges, and infrastructure needs.
Agentic AI autonomously plans, executes, and adapts strategies to reach goals, going beyond simple prompt-response.
Advances in language models, tool integration, and long context windows are making agentic AI practical for complex workflows.
In mission-critical scenarios like disaster inspection and search-and-rescue, communication-limited robots must make reliable onboard decisions. Episodic memory reuse, though low-cost, can be unsafe due to changed topology or insufficient resources, leading to 'memory traps'. This paper presents MemoGuard, a lightweight adaptive runtime that validates memories against topology, resource, and outcome contracts before reuse, invoking fallback only when validation fails. In a corridor-inspection simulator, MemoGuard reduces battery safety violations by 76.6% over similarity-only top-1 reuse and reduces fallback calls by 21.4% over always reasoning. On an NVIDIA Jetson AGX Xavier with local llama3.2:3b fallback, it avoids 3.67 s and 36.97 J overhead per trial.
Introduces 'memory traps': high-similarity but execution-invalid episodic memories.
MemoGuard validates memories via topology, resource, and outcome contracts before reuse.
A Partial Information Decomposition (PID) framework is used to select the most informative MRI contrast pair prior to training, reducing computational cost for multi-contrast 3D brain tumor segmentation. Applied to T1n, T1c, T2w, and T2-FLAIR, it selected T1c+T2-FLAIR, which achieved mean Dice 0.676 vs 0.687 for all four inputs on lightweight 3D U-Nets.
PID framework ranks input pairs by redundant, unique, and synergistic information about tumor burden
T1c+T2-FLAIR selected as best two-input configuration, second only to full four-input set
The article details how Silicon Valley's embrace of AI is leading to widespread layoffs and a sense of precarity among tech workers, who once enjoyed immense job security and high salaries. Through personal stories and expert analysis, it illustrates the shift toward automation and its impact on the workforce.
A former Meta executive recounts losing her job after four rounds of layoffs and struggling to find new opportunities.
Tech companies are ranking employees based on AI usage, creating a cutthroat environment.
A single 24GB GPU is the practical floor for serious local inference. This guide compares six open-weight models that fit one card at Q4_K_M, including Qwen3.6, Gemma 4, Mistral Small, gpt-oss-20b, and DeepSeek-R1-Distill. It covers VRAM fit, licensing, and the job each does best.
24GB is the practical floor: run right-sized 20B–35B models, not the biggest 70B quant you can squeeze in.
Qwen3.6-27B is the strongest all-around default; DeepSeek-R1-Distill-Qwen-32B is the tightest fit at ~18–20GB.
Headroom is a 30-second browser-based test that measures your GPU's memory bandwidth ceiling for local AI inference, requiring no model download, using synthetic weights, and needing WebGPU support. It measures bandwidth via three independent methods and detects a subgroup bug that can cause silent garbage output in browser LLMs. Validations show in-browser engines are launch-overhead-bound, with 2-3x hardware headroom.
30-second browser test, no model download, uses synthetic weights
Three independent bandwidth measurements: buffer copy, triad, and pure-read reduction
Together AI and Y Combinator have partnered to provide YC startups with a dedicated GPU cluster, addressing the compute bottleneck. Startups can commit for weeks instead of two-year contracts, and manage GPUs directly through a self-service portal without YC involvement. The cluster is fully utilized and supports needs from single-node to large-scale expansion.
Together AI and YC launch first dedicated GPU cluster for YC startups, providing direct compute access.
Commitments are for weeks instead of two-year contracts, freeing capital for other business needs.
A curated collection of free resources for AI researchers, including over $1000 in free compute credits, research guides, community forums, and more to help students start AI research without spending a dollar.
In an exclusive interview with CNBC, TSMC CFO Wendell Huang said the company is accelerating its Arizona factory expansion to meet multi-year structural AI demand. An additional $100 billion investment raises total U.S. commitment to $265 billion, with full-year capex revised up to $60-64 billion. TSMC is converting 5nm capacity to 3nm, and 2nm technology will drive next quarter's revenue.
TSMC commits an additional $100 billion to Arizona fab, bringing total investment to $265 billion.
The company is converting 5nm capacity to advanced 3nm nodes to meet AI demand.
Feyn Labs has released SQRL, a family of text-to-SQL models that inspect a database with read-only probes before committing to a query. The flagship SQRL-35B-A3B reports 70.6% execution accuracy on BIRD Dev, edging Claude Opus 4.6, and distills into self-hostable 4B and 9B checkpoints.
SQRL runs read-only probes to inspect the database before writing a final query.
SQRL-35B-A3B achieves 70.6% execution accuracy on BIRD Dev, surpassing Claude Opus 4.6 at 68.77%.
Alibaba's Qwen team previewed Qwen3.8-Max-Preview, a 2.4 trillion-parameter multimodal MoE model it calls "second only to Fable 5." The preview is live on Token Plan, Qoder, and QoderWork at 10% of standard pricing. What is not live: any benchmark table, model card, license, per-token price, or active-parameter count. This breakdown separates what Alibaba confirmed from what it only claimed.
Qwen3.8-Max-Preview is live via Token Plan, Qoder, and QoderWork at 10% of standard pricing.
The 2.4T parameter count and "second only to Fable 5" ranking are Alibaba's claims, not verified benchmarks.
A Forbes article profiles luxury escorts in Silicon Valley who are paid for their knowledge of AI and crypto. However, this is not a new phenomenon—similar roles have existed for centuries in the form of geishas, courtesans, and salon hostesses. The article argues that the human desire for intellectual companionship long predates AI.
Silicon Valley escorts now offer conversation on AGI and GPUs, charging thousands per meeting.
Historical parallels include Japanese geishas, Greek hetaerae, and French courtesans.
This week's developments in AI shift focus from raw scale to distribution and openness. Highlights include Thinking Machines' open-weight Inkling, Moonshot AI's 2.8T parameter Kimi K3, PrismML's phone-runnable Bonsai 27B, OpenAI's self-play red-teaming system GPT-Red, and Xi Jinping's call for open-source AI as a global public good at the World AI Conference in Shanghai.
Thinking Machines released Inkling, a 975B MoE model with open weights and 1M context
Moonshot AI unveiled Kimi K3, a 2.8T parameter model optimized for long-horizon tasks
Artificial intelligence is evolving from rule-based programming (Software 1.0) to learned models (Software 2.0) and now to natural language interfaces with emergent reasoning (Software 3.0). LLMs are a general-purpose technology with broad impact, but they lack physical grounding. The next frontier is spatial intelligence and physical AI, driving autonomous vehicles and humanoids.
AI has transitioned from human-coded logic to machine learning and now to natural language interfaces.
LLMs are a general-purpose technology, similar to electricity, with broad applicability.
Zlvox is a platform offering 25 free developer tools covering AI, JSON, PDF, image processing, and more, with a strong emphasis on privacy (client-side processing, zero data retention) and performance. No sign-up required.
25 free developer tools covering AI, JSON, PDF, images, and more, with zero tracking.
All tools run client-side; data never leaves the browser and is not stored.
Three Chinese labs' flagship open-weight MoE models—Kimi K3, DeepSeek V4 Pro, and GLM-5.2—each excel in benchmarks, licensing, and cost. Kimi K3 leads in capability but is API-only; DeepSeek V4 Pro is cheapest and fully open; GLM-5.2 balances speed and deployability.
Kimi K3 (2.8T params) tops the AAI Index at ~57 but weights won't be available until July 27 under a Modified MIT license.
DeepSeek V4 Pro (1.6T params) is MIT-licensed, costs ~$0.04 per task, and offers immediate open weights.
This tutorial provides a step-by-step guide to fine-tune Qwen3-0.6B with LoRA using NVIDIA NeMo AutoModel on a single GPU in Google Colab. It covers environment setup, recipe patching, training, evaluation, and Python API usage.
Set up NeMo AutoModel environment on Colab single GPU
Load and adjust Qwen3-0.6B LoRA fine-tuning recipe
New York City mayor Zohran Mamdani released a 'Rental Ripoff Report' recommending that landlords and realtors disclose the use of AI in altering rental listings, including images. The crackdown follows thousands of tenant complaints and aims to combat deceptive practices, supporting tenant unions and expanding bargaining rights.
Mayor Mamdani's report requires landlords to disclose AI-altered property images.
AI-generated deceptive images are increasingly used in real estate listings.
NVIDIA DeepStream 9.1 introduces 13 agentic skills that let coding agents like Claude Code and Codex build multi-camera video analytics pipelines from natural-language prompts. Multi-View 3D Tracking (MV3DT) fuses per-camera detections into one shared 3D world with a globally consistent object ID, while AutoMagicCalib (AMC) removes manual camera calibration. The release also adds JetPack 7.2 support and a unified open-source GitHub monorepo.
Talon is a multi-platform, self-hosted AI agent framework supporting Telegram, Discord, Microsoft Teams, terminal, and a cross-platform desktop/mobile app. It offers pluggable backends (Claude Agent SDK, Kilo, OpenCode, Codex, OpenAI Agents) and full MCP tool access, with background agents, goal management, skill system, event bus, and hot-reloadable plugins. The architecture is clean, with frontend and backend independent, making it highly extensible.
Supports multiple frontends (Telegram, Discord, Teams, terminal, desktop/mobile) and backends (Claude, Kilo, OpenCode, Codex, OpenAI Agents) with rich MCP tools.
Features background agents (heartbeat, dream), persistent goals, skill system (SKILL.md), and triggers for proactive task advancement.
The AI boom is increasingly financed by debt, but investor demand is falling as hyperscalers accelerate bond issuance. Amazon's recent bond sale required higher yields due to lower demand, with order coverage dropping. AI bond supply is surging while investors demand wider spreads. Meanwhile, the breakthrough performance of Chinese AI model Kimi K3 raises concerns about the sustainability of US AI spending, potentially leading to an economic slowdown.
Since early 2025, Alphabet, Meta, Amazon, and Oracle have issued over $300 billion in bonds.
Investor demand for AI bonds is declining; Amazon's bond orders fell from 3.2x to 2.5x coverage.
PixelUp is a lightweight AI video upscaler for Windows that runs entirely offline. It uses FSRCNN and ESPCN deep learning models to upscale low-resolution videos to high-definition quality with fast processing and hardware acceleration. The software offers one-time purchase pricing ($19) with no subscriptions, batch processing, lossless audio syncing, and full privacy protection.
100% local processing, no internet or account required, ensuring data privacy.
Utilizes optimized FSRCNN and ESPCN models for rapid detail reconstruction.
AIpine is an iPhone app for viewing, previewing, and organizing AI-generated files like JSX, HTML, Mermaid diagrams, SVGs, and more, fully offline with no account or cloud required.
AIpine provides preview and source code inspection for various AI-generated file formats.
The app works fully offline, storing files locally without requiring accounts or cloud services.