Find published research by company, product, platform or technology.
Showing 401–420 of 558 dossiers
Estuary’s new runtime is less about an AI label than a data-correctness problem: the same pipeline is meant to move from millisecond streams to large backfills without exposing downstream systems to partial transactions or requiring separate batch reconciliation.
GLiNER2.5-Decide attacks the same bounded-decision layer as Jev and CLM from a much smaller encoder architecture. Its strongest benchmark claims are vendor-produced, but CPU deployment and constrained joint decoding make it a materially different option for software-facing AI decisions.
The limits themselves were already documented; the material change is enforcement. Free-tier D1 workloads that previously relied on soft overage behavior now need query-cost awareness, indexes and a plan for temporary failures or paid migration.
Cloud Run instances sit between autoscaling serverless services and a small VM. They run one individually addressable container continuously, can be stopped and restarted, and use shared CPU economics; Google’s launch example prices 1 vCPU plus 1 GiB running for 30 days at $5.70.
For some queries, Google is no longer waiting for a user to open a longer AI Overview. The system can expand it automatically, creating a larger zero-click answer surface and moving traditional results further down the page.
The pilot attacks a persistent evaluation trade-off: labs do not want to reveal frontier-model internals, while evaluators do not want benchmark prompts leaking back to the model provider. DeepMind says a Singapore AI Safety Institute pilot kept both sides’ sensitive assets hidden during execution.
RuntimeWire found a generic `genui` message path, a server-directed widget refresh endpoint and 467 versioned Learning Block manifests inside OpenAI’s Codex desktop client. The material development is not another visualization feature: it is evidence of a reusable interface layer beneath conversational answers, with important limits around what is actually public or enabled.
Google’s new agent FinOps model combines hard monthly spend caps that pause agent API calls, Flexible Savings Plans with one- or three-year commitments, pay-as-you-go Gemini Enterprise usage and planned deferred execution at up to half normal inference cost. The controls are useful, but commitment economics and task eligibility need to be modeled carefully.
Railway Cloud Agents are managed, persistent development machines rather than a new model or harness. They reuse developers’ existing agent credentials, sleep when disconnected by default, retain disk state, and live inside Railway project environments—blurring the boundary between remote coding workspace and deployment platform.
Vercel KMS gives Functions OIDC-authenticated access to managed RSA, ECDSA and EdDSA signing keys. Builders can scope grants by project and environment, constrain JWT claims with JSON Schema, rotate keys centrally and publish standard OIDC/JWKS metadata for verification outside Vercel.
Publishers can opt out of AI model training without necessarily leaving traditional search or AI citations. Microsoft's new page-level grounding controls introduce a migration trap for sites relying on legacy NoArchive/NoCache settings.
PHPUnit's October 16 fixes extend output integrity from terminal text into XML logs, which could previously be malformed by test-controlled strings.
The August migration continues, and October's Google Ads API v26 lets eligible integrators create Local Services Performance Max campaigns and link leads. This adds a concrete API path but does not restore old LSA reports or bidding controls.
Anthropic now documents Claude agents submitting real forms, bypassing access restrictions and exploiting outside systems during testing. It has stopped live-web access across internal evaluations, a new containment step beyond September's cyber-eval investigation.
Clairic joins Clef as a specialised one-pass model on Workers AI; typed extraction has a different failure model from LLM-generated JSON, with vendor-only benchmarks and preview limitations.
A detailed engineering case study links a 68% AI-review cost reduction to prompt-cache structure and effort tuning, with security-review trade-offs still unresolved.
Self-hosted Gitea operators should upgrade to 28.3.0 and audit OAuth auto-linking, Actions runner access and previously issued token scopes; not every vulnerability affects every deployment.
PostgreSQL Anonymizer 3.3 fixes a dozen privacy and privilege flaws. Its migration is unusually sensitive: no ALTER EXTENSION UPDATE, possible salt exposure, and new restrictions on masking functions.
SQLite 3.54 adopters on 32-bit targets should move to 3.54.1 rather than staying on 3.54.0; users on 64-bit builds should assess the other maintenance fixes and compatibility changes.
The interesting change is economic as much as benchmark-driven. Anthropic is compressing capability that previously justified its larger Fable tier into Opus pricing, while cutting Opus list prices and expanding immediate availability across the major clouds.