The archive

Latest articles

Reporting, explainers, and analysis on data and AI, updated continuously by the newsroom.

AI peer review line showing combined ICLR, ICML, and NeurIPS submissions rising from 17,051 in 2020 to an estimated 73,883 in 2026.

AI peer review gets a serious scale test at ICML

AI peer review is moving into production workflow. Google's PAT found 89.7% of tested math errors, but review power remains human.

Lars Cornelissen · Jun 29, 2026

AI coding agent malware chain with 3 hidden indirection steps and 1 runtime DNS payload outside the GitHub repo.

AI coding agent malware hides in clean GitHub repos

AI coding agent malware can hide behind a clean repo. Lock down setup execution, DNS egress, and agent permissions before rollout.

Lars Cornelissen · Jun 29, 2026

VS Code Tasks supply chain attack affected 2 npm packages and 16 Go packages, according to JFrog Security Research.

VS Code Tasks supply chain attack needs new checks

VS Code Tasks supply chain attack is a package hijack pattern that can run outside npm lifecycle scripts, so scan editor configs now.

Lars Cornelissen · Jun 29, 2026

Unity AI Gateway budgets example showing a $5,000 shared Genie budget, $100 per-user threshold, $200 genie-code override, and $300 power-users override.

Unity AI Gateway budgets: the spend guardrail guide

Unity AI Gateway budgets set shared and per-user AI spend limits for Genie and gateway traffic, with alerts and blocking for admins.

Lars Cornelissen · Jun 29, 2026

LineShine supercomputer at 2.198 exaflops versus El Capitan at 1.809, Frontier at 1.353, Aurora at 1.012, and JUPITER at 1.000.

LineShine supercomputer makes CPU scale political again

LineShine supercomputer is a 2.198 exaflop CPU machine. Treat it as a supply chain signal, with a power bill attached.

Lars Cornelissen · Jun 29, 2026

Matillion Maia BigQuery support covers 29 transformation components, 40 orchestration components, and 122 connectors on Google BigQuery.

Matillion Maia BigQuery support: setup, cost, limits

Matillion Maia BigQuery support brings GCP warehouses into Maia on Current now, with Stable expected August 1, 2026.

Lars Cornelissen · Jun 29, 2026

Dynamic Iceberg table replication changed from skipped before general availability to supported at general availability, encoded as 0 before and 1 after.

Dynamic Iceberg table replication, costed honestly

Dynamic Iceberg table replication now keeps Snowflake managed Iceberg pipelines in DR plans, but refresh and transfer costs move to the target account.

Lars Cornelissen · Jun 29, 2026

Apple AI price hike raises HomePod mini by 30.3 percent, iPad Air by 25.0 percent, iPad Pro by 20.0 percent, MacBook Air by 18.2 percent, MacBook Pro by 17.7 percent, and MacBook Neo by 16.7 percent.

Apple AI price hike makes memory a consumer tax now

Apple AI price hike is a hardware margin test: memory costs are rising, but Apple's $29.6B profit makes the pass-through harder to defend.

Lars Cornelissen · Jun 28, 2026

Apple CXMT memory deal chart showing TrendForce midpoint price increases: conventional DRAM at 57.5 percent in Q1 2026 and 60.5 percent in Q2 2026, NAND at 35.5 percent and 72.5 percent.

Apple CXMT memory deal makes RAM a policy fight now

Apple CXMT memory deal is a supply chain stress test: DRAM prices are up 58 to 63 percent this quarter, making policy risk a BOM risk.

Lars Cornelissen · Jun 28, 2026

Cisco Unified CM CVE timeline showing 22 days from Cisco's June 3 advisory to CISA's June 25 KEV addition, then 3 days to the June 28 deadline.

Cisco Unified CM CVE gets a weekend patch clock

Cisco Unified CM CVE is now a patch-or-isolate job: CISA set a June 28 deadline after active exploitation of CVE-2026-20230.

Lars Cornelissen · Jun 27, 2026

Timeline for Signal backup recovery keys: Secure Backups announced at day 0, FBI and CISA messaging-app phishing warning at day 193, and CISA June 26 update at day 291.

Signal backup recovery keys become the weak link for ops

Signal backup recovery keys can expose historical chats after one phishing win. Treat messenger backups like identity infrastructure now.

Lars Cornelissen · Jun 27, 2026

Scatter of about 17 dictionary concepts plotted by cosine distance from a seed word, clustered into a near band of banned cliches and a middle band of surprising-but-related lenses the generator is allowed to use.

How to make an LLM escape its own prior with geometry

Novel Search Space breaks an LLM out of its prior by ranking 80,000 dictionary words by embedding distance, banning the obvious neighbours, and forcing the model to brainstorm only from a surprising-but-related band.

Lars Cornelissen · Jun 27, 2026