@infoq.com#AI tokens got 1,000% cheaper. Enterprise AI spend rose 222%. Why?
Erik Peterson breaks down the hidden economics of AI engineering & the road to autonomous cloud by 2028:
➡️ The tokenmaxxing myth
➡️ Cost-to-build vs. cost-to-serve
➡️ Unit economics over vanity KPIs
@infoq.comAI doubled Spotify’s code output.
Monthly merged code changes jumped from 8K to 17K. Now, Spotify is rethinking testing, rollback safeguards, deployment timing, observability, and workload prioritization to keep delivery stable.
@infoq.com#AWS enhances Amazon EventBridge with shared event buses across accounts, ordered delivery, event replay, content-based deduplication, and simplified subscriber management.
The update also introduces publisher/subscriber-based pricing.
🔗 Learn more: buff.ly/2uHeekv
@infoq.comChoosing the wrong data layout can hurt lakehouse performance.
This #InfoQ article compares Liquid Clustering, Z-Ordering, and partitioning across 4 Delta Lake workloads, uncovering 2 undocumented failure modes that could already be affecting production systems.
@infoq.comIs mandatory #CodeReview DEAD in the age of #AI?
2-week sprints feel too long. Reviewing every line of code isn’t sustainable.
Hannah Foxwell explains why software teams need to rethink peer review - and bring the joy back to engineering.
@infoq.comFirst-time manager? Start with the skills that matter most.
🎙️ Matt Gjertsen, founder of Built and author of Minimum Viable Manager, shares how to build trust, give effective feedback, and lead people in the age of AI.
🎧 Listen now: buff.ly/7chqWwx
@infoq.comCloudflare’s new Traces open beta adds OpenTelemetry spans for security rules, transformations, caching, routing, and origin handling, with W3C trace context propagation and targeted sampling through Trace Rules.
🔗 Learn more: buff.ly/4mEb1Y1
@infoq.comWhat if #LLMs could manage & edit their own context?
Context Language Models (CLMs) allow AI to update their context like a file - rewriting past notes, pruning fluff, and boosting accuracy by up to 35% while drastically cutting compute costs.
@infoq.comShopify rebuilt its Checkout Blocks app with Preact and Polaris web components, cutting bundle sizes by up to 85% and median load times by ~8%.
The key? Replacing React with Preact and enforcing a strict 64KB gzip limit per extension.
📖 Read now: buff.ly/NwOATfr
@infoq.comYour GPUs are fast. Your data pipeline isn’t.
Every ML training run pays a data movement tax. Vortex bypasses the CPU bottleneck to stream compressed data directly to GPU memory.
🎥 Onur Satici explains how it works! Watch now: buff.ly/BbBbLq5
@infoq.comMicrosoft releases #Aspire 13.6 with dashboard updates, Java and Rust hosting packages, and new CLI options.
➡️ SQLite-backed telemetry
➡️ Up to 10 completed runs per app
➡️ Portable volume paths
➡️ Breaking changes to MongoDB TLS and the Cosmos DB emulator
📖 Learn more: buff.ly/jASyCDl
@infoq.com#Cloudflare recently launched the open beta of cf - an open-source CLI built for developers & AI agents.
Written in TypeScript, it unifies access to Cloudflare services with structured output and command discovery.
📖 Learn more: buff.ly/ZMBPPwm
@infoq.comAI coding agents are moving faster than product teams can keep up. What needs to change?
Hannah Foxwell shares 3 anchors for navigating AI-driven velocity:
🎯 Build something worth building
🛡️ Speed requires safety
👥 People matter
➡️ Watch now: buff.ly/hlLsuSv
@infoq.com#AICodingAgents write code faster. But 79% of engineering leaders say overall release speed hasn't improved.
Why the gap?
➡️ 16.9 hours/week spent debugging AI-generated code
➡️ 35% of AI code reaches production before teams fully understand it
📖 Read more: buff.ly/RoDA9iQ
@infoq.comGrab migrated its high-volume Counter Service to #Aerospike, cutting p99 read latency by ~50%, disk usage from 3 TB to 1 TB, and cost per node by 45–50%.
How? Storage abstraction, shadow traffic, data parity checks, gradual migration, and a redesigned data model.
@infoq.comHow do you guarantee session-level message ordering across thousands of #ApacheKafka channels?
Joshua Oluikpe explains a Go-based approach using consistent hashing, retries, and contiguous watermark commits.
🔍 Dive deep - buff.ly/0z4SKkI
@m-kumagai.bsky.social【お詫び】【再発防止について】
セキュリティ事業を担うグループとして、今回のinfoQの情報漏えい事故を重く受け止めております。
お客様をはじめ、関係する皆様に多大なご迷惑とご心配をおかけしましたことを、深くお詫び申し上げます。
不正侵入を防げなかった事実を厳粛に受け止め、グループ全社・全サービスの緊急再点検と、8,300人のパートナーへの再教育を実施し、再発防止に全力で取り組んでまいります。
@infoq.comIt’s been a busy week in the #Java ecosystem!
➤ The GA releases of JobRunr 9.0 and OpenXava 8.0; point releases for Quarkus, Micronaut, LangChain4j, Eclipse JNoSQL; ADK for Java and ADK for Kotlin; and introducing Lathe, a new Java language server.
🔗 #InfoQ News Roundup: buff.ly/UWHCgl4
@infoq.comSecurity is moving from human vigilance to automated, system-level defenses.
Explore:
→ CHERI hardware memory safety for legacy C/C++
→ SBOMs under the EU Cyber Resilience Act
→ Crypto-agility for the post-quantum era
→ Least-privilege access for AI agents
@infoq.comOpenTelemetry’s Kubernetes Attributes Processor is now stable in v1.0.0!
⚠️ K8s Semantic Conventions introduce breaking attribute name changes. Update dashboards and alerts when migrating.
Read the full breakdown 👉 buff.ly/IIkHNCN
