Hacker News

Article URL: https://herecomesthemoon.net/2025/11/how-many-options-fit-into-a-boolean/

Comments URL: https://news.ycombinator.com/item?id=47284781

Points: 1

# Comments: 0

Article URL: https://apnews.com/article/georgia-electric-vehicle-battery-manufacturing-layoffs-workers-79a4ec7a5b7f2816f64033f8b3d4898d

Comments URL: https://news.ycombinator.com/item?id=47284778

Points: 1

# Comments: 0

Article URL: https://www.ft.com/content/eaecc3ce-ba4e-4e19-9d0a-6cf1536a5aed

Comments URL: https://news.ycombinator.com/item?id=47284766

Points: 2

# Comments: 0

Article URL: https://cat.aadishv.dev/

Comments URL: https://news.ycombinator.com/item?id=47284763

Points: 1

# Comments: 1

Article URL: https://flowingdata.com/2026/03/05/unemployment-reasons/

Comments URL: https://news.ycombinator.com/item?id=47284752

Points: 2

# Comments: 0

Article URL: https://kerkour.com/rust-postgres-everything

Comments URL: https://news.ycombinator.com/item?id=47284748

Points: 2

# Comments: 0

Article URL: https://github.com/Naveenub/quantum-pulse

Comments URL: https://news.ycombinator.com/item?id=47284742

Points: 1

# Comments: 0

Article URL: https://kelvinfichter.com/pages/thoughts/train-harder/

Comments URL: https://news.ycombinator.com/item?id=47284719

Points: 1

# Comments: 0

I was paying $100/mo for Claude Code Pro and had no idea where it was going. I'd hit the 5-hour rate limit constantly, but couldn't tell which sessions were burning through my allocation or whether Opus was worth the premium over Sonnet for my workflows. So I built a tool to find out.

BurnRate is a local CLI that parses your AI coding tool session data and gives you a full cost analytics dashboard. It tracks Claude Code, Cursor IDE, and OpenAI Codex in one place.

Everything runs 100% on your machine. Your session data, token counts, costs, prompts — none of it leaves your computer. No API key to paste, no telemetry. It reads the local session files your tools already generate and does the math.

Out of the box you get: multi-provider cost tracking, 10 different analytics views (daily trends, per-session breakdown, model usage split, token efficiency), an optimization engine with 23 rules that suggests concrete config changes to reduce spend, usage limit monitoring so you know when you're approaching rate limits, side-by-side provider comparison, and budget alerts when you're on track to blow past a monthly cap.

For managers, there's a team dashboard where devs can optionally push anonymized usage snapshots to a shared view. Useful for understanding team-wide AI tool costs and figuring out which plans actually make sense. Free tier available, Pro is $9/mo, Team is $29/mo.Happy to answer any questions. Feedback welcome.

Comments URL: https://news.ycombinator.com/item?id=47284701

Points: 1

# Comments: 0

Article URL: https://iisc.ac.in/events/worming-out-molecular-secrets-behind-collective-behaviour/

Comments URL: https://news.ycombinator.com/item?id=47284699

Points: 2

# Comments: 0

Article URL: https://github.com/shubhdeep12/resend-cli

Comments URL: https://news.ycombinator.com/item?id=47284698

Points: 1

# Comments: 0

Article URL: https://appmakes.github.io/Rai/

Comments URL: https://news.ycombinator.com/item?id=47284697

Points: 1

# Comments: 0

Article URL: https://github.com/fivefilters/ftr-site-config

Comments URL: https://news.ycombinator.com/item?id=47284694

Points: 1

# Comments: 0

Article URL: https://www.universetoday.com/articles/astronomers-using-meerkat-spot-a-cosmic-laser-halfway-across-the-universe

Comments URL: https://news.ycombinator.com/item?id=47284689

Points: 2

# Comments: 0

Article URL: https://www.nbcnews.com/politics/white-house/trump-privately-shown-serious-interest-us-ground-troops-iran-rcna262176

Comments URL: https://news.ycombinator.com/item?id=47284678

Points: 4

# Comments: 2

Article URL: https://arxiv.org/abs/2603.01896

Comments URL: https://news.ycombinator.com/item?id=47284670

Points: 1

# Comments: 0

Hi HN,

I’ve been experimenting with running small language models directly on mobile devices and built a small Android app called EdgeDox.

The idea was to make document AI usable without sending files to a cloud service. Many existing tools require uploading PDFs or documents to a server, which can be a privacy concern.

EdgeDox runs a lightweight language model (Qwen3.5-0.8B) locally on the device so documents stay on the phone.

Current features:

• Ask questions about PDFs • Document summarization • Extract key points from long documents • Works completely offline • No accounts or server processing

The model runs locally using mobile inference (MNN). I'm experimenting with quantized models and other optimizations to keep memory usage and latency reasonable on mid-range Android devices.

Some challenges so far:

• balancing context size with mobile memory limits • improving response latency on CPU-only devices • reducing model load time

The project is still in early beta, and I’m mainly looking for feedback from people experimenting with on-device AI or mobile inference.

Play Store: https://play.google.com/store/apps/details?id=io.cyberfly.ed...

Comments URL: https://news.ycombinator.com/item?id=47284668

Points: 1

# Comments: 0

Article URL: https://www.gamingonlinux.com/2026/03/ea-javelin-anticheat-job-listing-mentions-future-support-for-linux-and-proton/

Comments URL: https://news.ycombinator.com/item?id=47284621

Points: 2

# Comments: 1

AI agents are increasingly performing automated web research — browsing pages, following links, and sometimes clicking results as part of information gathering.

There's a small but potentially significant side effect: these systems can end up clicking paid advertisements.

Most online advertising runs on a pay-per-click (PPC) model. When a human clicks an ad, there's at least some level of commercial intent. When an AI agent clicks an ad during automated research, there's zero purchase intent — but the advertiser may still be charged.

At the individual level this is negligible. But AI agents are beginning to operate at scale — millions of automated queries. The cumulative effect on advertisers, particularly small businesses with tight budgets, could become meaningful.

This raises a few questions:

1. Should AI agents avoid clicking sponsored/promoted results by default? 2. Should browsers and agent frameworks detect labels like "Sponsored," "Promoted," or "Ad" and skip those results unless explicitly opted in?

Secondary effects worth considering: unintended ad spend for advertisers, distortion of click-through analytics, and reduced research quality (ad placement reflects budget more than relevance).

The web's ad-funded model depends on clicks having some commercial signal. If AI agents start generating ad clicks at scale with no purchase intent, it could quietly distort that ecosystem.

Curious how engineers and AI developers here think about this — both from an agent design standpoint and from the web economics angle.

Comments URL: https://news.ycombinator.com/item?id=47284593

Points: 2

# Comments: 2

Article URL: https://serd.es/2026/03/06/TCXO-failure-analysis.html

Comments URL: https://news.ycombinator.com/item?id=47284587

Points: 1

# Comments: 0