Topic
Everything filed under AI, newest first.
RSS · JSON · All topics
Policy and Disclosure: 2026 Edition
Project Zero is replacing its "90 + 30" disclosure policy: citing AI-accelerated vulnerability discovery and exploitation, it will disclose vulnerabilities 90 days after reporting regardless of whether a patch has shipped, and will no longer make fix status a decision point.
5 min · 1,129 words
Weight-Space Programming with DSPy
Drew Breunig introduces dspy-wsp, a prototype DSPy optimizer that lets a frontier model observe and directly edit a small model's weights, without gradients or training, to improve its benchmark performance.
4 min · 824 words
Beating the frontier on NetHack
Antithesis explains how its exploration fuzzer, which can now find a path to victory within five wall-clock minutes on nearly 80% of NetHack seeds, got there, and how the same search-and-optimization approach applies to testing software.
77 min · 17,750 words
Introducing Edge0-145B: Frontier AI Capabilities on Your Own Computer
Edge0 open-sources Edge0-145B and Edge0-120B, which it describes as the first open-source on-device models at 145B parameters, along with a unified inference framework and deployment tools for running coding, reasoning and agent workloads locally without cloud token fees.
15 min · 3,432 words
Erik Hoel proposes qualiaphobic AI: neural architectures deliberately engineered so no theory of consciousness could deem them conscious, via lookup-table anchoring, fragmenting phenomenology, and adversarial design, arguing this solves AI welfare and keeps humans the only moral patients.
6 min · 1,337 words
“…and the proof is checked in Lean”
Kevin Buzzard explains how to interpret claims that a theorem is correct because it was checked in Lean, and which tests are needed when the Lean code is AI-generated, citing OpenAI's recent theorem dump.
8 min · 1,864 words
The curious case of the north-east arrow
Arsh Kohli samples about 4,000 new public GitHub web repos and AI-attributed commits to show the north-east arrow (↗) spreading across LLM-built landing pages, from roughly 1 in 20 repos early in 2026 to about 1 in 6 since September, with OpenAI's Codex era models the heaviest users.
4 min · 924 words
Introducing Claude Meetings, now in beta
Anthropic launches Claude Meetings in beta on Team and Enterprise plans: Claude joins video calls as a visible participant, answers questions, uses connected tools with permission, writes notes and action items, and builds docs and whiteboards live during the meeting.
3 min · 702 words
Is there a ‘right amount’ of AI?
Watts Martin, an AI skeptic but not a purist, weighs how much LLM-generated code he can accept in the apps he uses, from Overcast to Fineliner, concluding that the less AI-generated code an app contains the easier it is to vet, and the least amount is always zero.
8 min · 1,853 words
Gemini 4 Hydrogen and Orchestrator
Google Antigravity rolls out Gemini 4 Hydrogen plus a new Orchestrator model option that dynamically hands work between Gemini 4 Carbon (planning) and Hydrogen (execution) within one trajectory, reaching near-Carbon success on Terminal-Bench 4.1 with 24% less quota and 19% faster completion.
3 min · 633 words
The Brave Search API now accepts agentic payments via the x402 protocol
Brave announces that its Search API now supports pay-as-you-go agentic payments via the open x402 protocol: agents pay per request in USDC on Base (about $0.005 per query) with no account or API key, using a 402 Payment Required challenge and signed payment retry.
2 min · 419 words
Top Chinese Frontier AI Models Carry the Fingerprint of Export-Controlled NVIDIA Blackwell GPUs
A forensic analysis of publicly released model checkpoints finds numerical patterns matching NVIDIA Blackwell hardware in six models released since April 2026 by DeepSeek, Moonshot AI, MiniMax and Ant Group, suggesting access to export-controlled GPUs; earlier models from the same labs match the older Hopper generation.
25 min · 5,671 words
Google introduces Gemini 4 Hydrogen, a fast, cost-efficient model in the Gemini 4 family built as a daily driver for everyday coding and reliable agentic workflows, with availability across the Gemini API, Gemini app, Search and Workspace.
8 min · 1,894 words
Attention to Mamba: A Recipe for Cross-Architecture Distillation
Apple ML Research (TMLR) proposes a two-stage recipe to distill a Transformer into an attention-free Mamba model via linearized attention, retaining most of Pythia-1B's performance (perplexity 14.01 vs 13.86).
1 min · 319 words
Which campus has the worst food? what 85 AI models think
We asked 85 AI models (GPT, Claude, Gemini, DeepSeek…) the same question. 7% picked UMass Amherst. See every answer and who dissented.
11 min · 2,476 words
My AI Agents Found an Unmapped Maya Site
A first-person account of pointing a team of AI agents at NASA lidar data and, in under 72 hours, surfacing a previously unreported Maya site in Mexico with a plaza, pyramid and platforms, then getting an official reply from the Mexican government.
6 min · 1,390 words
Fernando Borretti sketches a map of possible futures, laying out the different ways the arrival of advanced AI could play out for society and individuals.
7 min · 1,555 words
After 600 experiments, here's what I discovered about JEPA world models
Haddy Alchaer reports findings from 600 experiments with JEPA world models, showing how vanilla LeWM can be augmented with memory-graph subgoals to solve long-horizon physics planning tasks on out-of-distribution goals without extra training.
12 min · 2,720 words
The dawn of the Data Intelligence Engineer
MotherDuck's Jordan Tigani argues that AI agents now handle much of traditional data-team work, and proposes a new role, the Data Intelligence Engineer, who owns the data model, curates context, and makes sure the numbers agents produce are right.
10 min · 2,271 words
ExecuTorch WebGPU: PyTorch Models on the GPU, in the Browser
The PyTorch team introduces an ExecuTorch WebGPU backend that runs PyTorch models on the GPU in the web browser without converting to another framework: export with torch.export to a .pte file and run it via WebGPU.
4 min · 855 words