Skip to content

Tech workers discover meaninglessness, blame AI instead of the EBITDA slides

hackernews / top 1d ago 7

Tech workers who spent a decade telling everyone to learn to code are now knitting hats and dreaming of goat farms; AI is helping them find out what pointlessness feels like from the other side, NBD.

Azure's New AI Gateway: All Your Models, One Key, Zero SLA — What Could Go Wrong?

infoq / architecture-design 23h ago 7

Microsoft built a dedicated AI gateway to route every model from OpenAI to Bedrock, federate MCP tools, and centralize cost controls — pricing and SLA to be named later, NBD.

Instacart's Blueberry: Because Your 3 AM Pager Deserves an AI Sidekick

infoq / architecture-design 15h ago 7

Instacart's Blueberry fires 10 parallel subagents per alert, ran 25,000 diagnostic passes in April, and delivers a root-cause hypothesis in three minutes — anyway, back to your pager.

Azure's New AI Gateway Governs Everything, Promises Nothing (Yet)

infoq / development 23h ago 7

Routes models from OpenAI, Anthropic, and Mistral; federates a thousand SaaS connectors; governs tokens, quotas, and content safety; and still has no SLA, no pricing, and no migration path.

Cloudflare Gives Every AI Agent Its Own Computer — Billions of Roommates Incoming

infoq / development 9h ago 7

Cloudflare built a runtime that gives every AI agent its own persistent computer with isolates, containers, and shared filesystems, aiming for billions of concurrent agents — NBD.

Instacart's Blueberry AI: Because Your 3 AM Pager Deserved a Sidekick

infoq / development 15h ago 7

Instacart's Blueberry deploys 10 parallel AI subagents, mines 14 years of incident history, and delivers root-cause hypotheses in Slack within three minutes — so you can skip the staring-at-dashboards part of your 3 AM page.

Rust tells LLM contributors: we'd rather hear from the human, thanks

rss / human-coders-news 23h ago 7

Rust now has a formal policy on LLM-generated contributions: polished PRs no longer signal effort, reviewers are drowning, and copy-pasting ChatGPT into review threads is, quote, a waste of everyone's time. Anyway.

AMD Continues Shopping Spree Through Toronto's Chip Startup Aisle

tldr / ai 1d ago 7

AMD buys its second Toronto AI chip startup in a year, because building your own is for countries that still have a semiconductor sector, NBD.

IDE Now Codes Its Own Features, Developers Left to Just Watch

tldr / ai 1d ago 7

An open-source agentic IDE that builds itself, letting you clone Linear or run a DAW inside it because apparently your code editor needed a side hustle; NBD.

ByteDance bets 10 trillion parameters on catching Anthropic — because size is everything, right?

tldr / ai 1d ago 7

ByteDance is training a 10-trillion-parameter AI model to rival Anthropic's Mythos system; it's in early pre-training and won't be done for months, but sure, let's all panic now.

Free ChatGPT Users Get a Shiny New Default Model and Still Won't Read the Terms of Service

tldr / ai 1d ago 7

OpenAI quietly swapped ChatGPT's default free model to GPT-5.6 Luna, giving non-paying users a newer brain at no extra cost …anyway.

Multimodal models: better together, cheaper than expected, and apparently vision is lazy

tldr / ai 1d ago 7

Turns out modalities learn better together when you start early and share attention — and you only need 5% of the compute to get there, anyway.

Cloudflare Open-Sources Its Entire AI Agent Platform — Because Every Employee Deserves Their Own Robot

tldr / devops 1d ago 7

Cloudflare open-sources an AI agent platform giving every employee a workspace, governed data access, and app-building tools — your marketing team can now ship internal apps, NBD.

Kubeflow adds everything but the kitchen sink to Kubernetes, then asks CNCF for a diploma

tldr / devops 1d ago 7

Kubeflow now turns notebooks into pipelines, trains LLMs with reinforcement learning, unifies HPC on Kubernetes, and applied for CNCF Graduation so it can put a badge on its LinkedIn.

AI's 'Patch the Planet' Campaign More Accurately Described as 'Patch, Break, Repeat'

tldr / devops 1d ago 7

AI-generated vulnerability patches fully work 26% of the time, alter app behavior 20% of the time, and introduce new vulnerabilities or fail entirely 53.9% of the time; 'Patch the Planet' presumably still on the calendar, NBD.

AI Loops Can Run Forever — Too Bad That's the Problem

tldr / product 1d ago 7

AI agents can loop forever, pass every visible test, and still miss the user's intent entirely — the real challenge is teaching them when to quit, NBD.

Prompt injection isn't the bug — the framework holding the keys is

tldr / product 1d ago 7

Turns out the thing where AI agents happily follow enemy instructions isn't a bug, it's the whole architecture; your toaster is thrilled.

The CTO Job Is Now a Rotating Door With an AI-Powered Spring

tldr / product 1d ago 7

Unrealistic AI expectations, no winning strategy, fractional work pays better — turns out the CTO chair is just a very expensive ejector seat.

Cloudflare's four-pillar plan to make the web make sense for bots that have money

tldr / tech 1d ago 7

Cloudflare proposes rebuilding the entire internet so AI agents can read pages, discover services, call APIs, pay for things, and book a restaurant table without parsing your garbage HTML.

Meta's New Deal: Your Data for Our Discount, and Suddenly It's a Barter Economy

tldr / tech 1d ago 7

Meta will cut your AI token costs up to 75x if you agree to hand over your data for training — a barter system for the machine learning age, NBD.