The Archive

Trump administration delays announced AI executive order amid internal White House policy disagreements.

Listen to the session or watch below AI companies want to build systems that understand the external world and overcome the limitations of LLMs. Recent developments have brought world models to the forefront of the AI discussion. Watch a conversation with editor in chief Mat Honan, senior AI editor Will Douglas Heaven, and AI reporter…

Daytona CEO discusses agent infrastructure platform achieving 74% MoM growth, 850K daily runs, and bare metal sandboxes for RL evaluation.

User reports Qwen 3.6 35B enabling agentic workflows for DevOps, document processing, and code tasks via skill-chaining.

Reddit speculation about unreleased Google Gemini 3.5 Pro model; no concrete information provided.

Link: youtube.com

University graduates are booing and heckling corporate executives who praise AI during their commencement ceremonies, and the only people who seem to be genuinely surprised by this are the executives themselves. In a procession of viral videos, 2026 commencement speakers like former Google CEO Eric Schmidt face loud and sustained jeers from students after praising AI and describing the technology as both inevitable and mandatory. The videos have clearly struck a chord among young people entering a bleak job market in an increasingly unstable world. "They deserve everything they're getting," P...

Hivemind, an open-source Claude Code plugin that auto-generates reusable skills from repeated user prompts as slash commands.

You can actually make full manhwa story now. Characters stay same across panels, faces and feelings look right, and background also keep good. So far I make more than 20 pages, but I cannot upload all here, so I publish it in [https://www.vixal.art/en/explore/the-last-demon-king-s-son](https://www.vixal.art/en/explore/the-last-demon-king-s-son) I will keep working and try to finish

Qwen 3.7 open-weights model released; community discussion on LocalLLaMA highlights adoption momentum.

Simon Willison releases Datasette Agent, a conversational AI assistant for querying structured data with chart generation capabilities.

Google DeepMind launches accelerator program in Asia Pacific focused on environmental risk applications.

Spotify is partnering with Universal Music Group to let Premium subscribers create AI-generated song covers and remixes, with participating artists receiving a share of the revenue.

Google is about to look really different, and if you're not a fan of the AI overview feature, then you're not going to like what's coming.

Storytelling is core to humanity’s DNA, stemming from our impulse to express ideals, warnings, hopes, and experiences. Technology has always been woven through the medium and the distribution: from early humans’ innovation of natural pigments and charcoals for cave paintings to literal representation by the camera. The landscape of storytelling continues to shift under our…

Developer builds MCP-based agent loop using Claude Opus 4.7 to iteratively humanize AI-generated text via detector feedback.

LatitudeGames releases Equinox-31B, a Gemma-based 31B finetune blending dark adventure and slice-of-life storytelling data.

In quantitative finance, researchers build algorithms to trade assets, derivatives, and other financial instruments. A key part of that work is finding signals:... In quantitative finance, researchers build algorithms to trade assets, derivatives, and other financial instruments. A key part of that work is finding signals: patterns in messy market data that may help predict future returns. These signals can come from price and volume data, economic indicators, fundamentals, or alternative sources like news sentiment. For years… Source

datasette-agent-sprites 0.1a0 plugin enables Datasette Agent to execute commands in Fly Sprites sandbox environments.

Anthropic's June 15 Agent SDK pricing changes signal shift toward managed agents, pressuring third-party integrations and local deployment strategies.

Tokenisation is an integral part of the current NLP pipeline. Current tokenisation algorithms such as BPE and Unigram are greedy algorithms -- they make locally optimal decisions without considering the resulting vocabulary as a whole. We instead formulate tokeniser construction as a linear program and solve it using convex optimisation tools, yielding a new algorithm we call ConvexTok. We find ConvexTok consistently improves intrinsic tokenisation metrics and the bits-per-byte (BpB) achieved by language models; it also improves downstream task performance, but less consistently. Furthermore,...

We propose the Integrable Context-Dependent Demand Network (ICDN), a demand-first neural model for multiproduct retail demand. The model learns log-demand as a smooth, context-conditioned function of log-prices, allowing elasticities to be derived exactly from the learned demand surface. On the Dominick's beer dataset, ICDN improves out-of-sample generalization over a directed log-log benchmark and yields more stable, economically plausible elasticity estimates, especially for weakly identified cross-price effects.

Language models must now generalize out of the box to novel environments and work inside inference-scaling search procedures, such as AlphaEvolve, that select rollouts with a variety of task-specific reward functions. Unfortunately, the standard paradigm of LLM post-training optimizes a pre-specified scalar reward, often leading current LLMs to produce low-entropy response distributions and thus to struggle at displaying the diversity that inference-time search will require. We propose Vector Policy Optimization (VPO), an RL algorithm that explicitly trains policies to anticipate diverse down...

Anthropic launches 13 free certification courses on Claude and agentic AI; community skeptical about credential inflation on LinkedIn.

Exploration is a prerequisite for learning useful behaviors in sparse-reward, long-horizon tasks, particularly within 3D environments. Curiosity-driven reinforcement learning addresses this via intrinsic rewards derived from the mismatch between the agent's predictive model of the world and reality. However, translating this intrinsic motivation to complex, photorealistic environments remains difficult, as agents can become trapped in local loops and receive fresh rewards for revisiting forgotten states. In this work, we demonstrate that this failure stems from a lack of spatial persistence a...

Robustness, domain adaptation, photometric and occlusion invariance, compositional generalisation, temporal robustness, alignment safety, and classical anisotropic regularisation are usually treated as separate problems with separate method families. This paper argues that much of their shared structure is one statistical problem: estimate the covariance of label-preserving deployment nuisance, then regularise the encoder Jacobian along a matrix whose range covers that covariance (the matching principle). CORAL, adversarial training, IRM, augmentation, metric learning, Jacobian penalties, and...

We propose and analyze a conservative drifting method for one-step generative modeling. The method replaces the original displacement-based drifting velocity by a kernel density estimator (KDE)-gradient velocity, namely the difference of the kernel-smoothed data score and the kernel-smoothed model score. This velocity is a gradient field, addressing the non-conservatism issue identified for general displacement-based drifting fields. We prove continuous-time finite-particle convergence bounds for the conservative method on $\R^d$: a joint-entropy identity yields bounds for the empirical Stein...

Autonomous agentic systems are largely static after deployment: they do not learn from user interactions, and recurring failures persist until the next human-driven update ships a fix. Self-evolving agents have emerged in response, but all confine evolution to text-mutable artifacts -- skill files, prompt configurations, memory schemas, workflow graphs -- and leave the agent harness untouched. Since routing, hook ordering, state invariants, and dispatch live in code rather than in any text artifact, an entire class of structural failure is physically unreachable from the text layer. We argue ...

Linear attention replaces the unbounded cache of softmax attention with a fixed-size recurrent state, reducing sequence mixing to linear time and decoding to constant memory. The hard part is not just what to forget, but how to edit this compressed memory without scrambling existing associations. Delta-rule models subtract the current read before writing a new value, and Kimi Delta Attention (KDA) sharpens forgetting with channel-wise decay. But the active edit still uses a single scalar gate to control two different things: how much old content to erase on the key side and how much new conte...

Large language model (LLM)-based multi-agent systems increasingly rely on intermediate communication to coordinate complex tasks. While most existing systems communicate through natural language, recent work shows that latent communication, particularly through transformer key-value (KV) caches, can improve efficiency and preserve richer task-relevant information. However, KV caches also encode contextual inputs, intermediate reasoning states, and agent-specific information, creating an opaque channel through which sensitive content may propagate across agents without explicit textual disclos...