AI — Page 53

Boston startup says the 128K-context, 38T-token LFM2.5 upgrade delivers reliable agentic behavior, fast tool calling, and open weights for phones, laptops, PCs, robots, and lightweight servers.

A retweet from Peter Jansen amplified Noy Sternlicht's note that CHIMERA made ACL 2026's main conference; the arXiv preprint is live.

ICLR 2026 work recasts block-wise updates as reverse diffusion, reporting comparable results in vision, image generation, and language while storing activations for a single block.

The new connector makes image and video generation callable inside agent workflows and exposes models like Gen-4.5, Kling, and GPT image 2 right from chat.

Open-source and billed as the second piece of its full-stack suite, Genesis World 1.0 targets the 1x real-world speed bottleneck with a technical blog laying out the thesis.

Open-source 4B image models fit in 0.93GB and 1.21GB, and Bonsai Studio brings on-device generation to iPhone under Apache 2.0.

In a thread on X, Sushant Pandey announced Prava Pay, which the company says gives AI agents scoped, single-use Visa cards with passkey approvals so users can let bots buy things safely.

Mallery introduced SEND, offered a free month to commenters, and took aim at clunky dashboards and AI SDRs.

Founded by Berkeley and Stanford researchers, the data lab pays service workers to wear camera caps and sensors and says it already spans 100k+ contributors and 500 partners.

Aligned News cited a 300,000-star moment and a Google 'Spark' entrant; while unverified, the buzz spotlights OpenClaw's local, open, self-hosted agent thesis.

The ClickUp CEO says savings will fund million-dollar salary bands for AI-leveraged top performers, even as Gartner warns automation cuts do not guarantee returns.

A brief X post teased an agentic framework for research-level math, but shared no docs or team names identifying what AlphaProof Nexus is.

VP of Research Pushmeet Kohli points to a GitHub trove of Lean-formalized proofs and prose by AlphaProof Nexus, signaling progress while holding back the framework code.

A widely reshared ClaudeDevs note, surfaced by Aligned News in a post on X, says Auto Mode now runs on Pro with Sonnet 4.6 and Opus 4.7.

Haritz Puerto says a paper on decomposing and measuring evaluation awareness just dropped, plus a resource called EvalAwa..., but links and authorship were not shared.

In a new paper and open releases, Li and collaborators decompose evaluation awareness, test nine models across four benchmarks, and publish a factor-controlled dataset and code.

The tally counts only Hugging Face and Ollama pulls, hinting at on-device demand but leaving methodology and release timing unclear.

Coffee meetups land June 2nd, 4, and 6, with RSVPs running through Luma as Asimov convenes folks interested in humanoids and AI.

Demos show a Mac agent that records your cross-app workflow once, compiles it into a deterministic DSL, and replays it locally with zero recurring tokens. With freu-cli open-sourced and a local vision execution model coming, can ahead-of-time semantic compilation beat screenshot agents on cost and latency?

Socket researchers say the TrapDoor campaign planted credential-stealing payloads in more than 34 packages and 384+ versions, targeting crypto and AI developers.

Google DeepMind reports a full-featured AlphaProof Nexus agent solved 9 of 353 open Erdos problems at a few hundred dollars per problem and proved 44 of 492 OEIS conjectures; code and Lean proofs are on GitHub. This is an arXiv preprint and community validation is pending.

In a rare interview on The Knowledge Project, the OpenAI co-founder recounts quitting within hours, sketching a backup company, and why they stopped showing reasoning traces.

The app says it now runs a lightweight local LLM as an interrogator model and splits generation and editing into distinct modes for a cleaner workflow.

Aligned News flagged the move; Wikipedia lists Karpathy on Anthropic’s pretraining team in 2026, underscoring top-tier talent consolidation at frontier labs.