Apple releases iPhone Duo simulator and Xcode 27.1 beta (developer.apple.com ↗)
2 pointsby 2 minutes ago
1comments
The new AgentCore runtime: Elastic, optimized, and consistently fast starts (amazon.com ↗)
1 pointsby 3 minutes ago
—discuss
Love vs. Wiley Book Writers: A Publishing Scam Gets a Rare Comeuppance (writerbeware.blog ↗)
2 pointsby 3 minutes ago
—discuss
Record 38% of new vehicle sales to be EVs in 2030, study finds (news.harvard.edu ↗)
1 pointsby 4 minutes ago
—discuss
Putin Uses Unlicensed Windows (twitter.com/christogrozev ↗)
1 pointsby 5 minutes ago
—discuss
PC ports of old console games are the new AI vibe coding battleground (pcgamer.com ↗)
1 pointsby 6 minutes ago
—discuss
Classified F-35 stealth parts diverted to Hong Kong by shipper, gone missing (politico.com ↗)
2 pointsby 9 minutes ago
—discuss
Stratasys awarded 27.6M in damages in first of 2 cases against bambu lab (all3dp.com ↗)
2 pointsby 9 minutes ago
—discuss
Have You Seen Shannon (news.ycombinator.com ↗)
1 pointsby 10 minutes ago
—discuss
AI Lab building tycoon game (bedroomlabs.net ↗)
1 pointsby 10 minutes ago
—discuss
An Undercover Google Analyst Infiltrated a Notorious Supply-Chain Hacking Gang (wired.com ↗)
1 pointsby 12 minutes ago
—discuss
Before you build it, share it (coidea.co ↗)
1 pointsby 12 minutes ago
—discuss
I vibecoded a tool to shrink Golang codebase automatically (github.com/dhilst ↗)
1 pointsby 12 minutes ago
1comments
SSD-Llama: SSD-Native Inference for Trillion-Parameter Moe on a Consumer PC (arxiv.org ↗)
1 pointsby 14 minutes ago
—discuss
ChatGPT's Built-in Browser now supports Chrome extensions (twitter.com/jameszmsun ↗)
2 pointsby 14 minutes ago
—discuss
Keeping vLLM's Prefix Cache Warm Between Agent Turns (doug.sh ↗)
1 pointsby 15 minutes ago
—discuss
Near-real-time vision using DiffusionGemma in Jev mode (twitter.com/mmastrac ↗)
1 pointsby 15 minutes ago
—discuss
Anthropic finally adds AGENTS.md support to Claude Code (twitter.com/trq212 ↗)
3 pointsby 15 minutes ago
1comments
The Highway Tutorial (fastcode.substack.com ↗)
1 pointsby 16 minutes ago
1comments
Show HN: Iceoryx2 0.10: flatbuffer integration, zerocopy IPC with unbounded data (ekxide.io ↗)
2 pointsby 19 minutes ago
—discuss
Show HN: Open-source revenue recognition and analytics for Stripe and Metronome
2 pointsby 21 minutes ago
—discuss
I built a project database for data science agents (yizhao95.github.io ↗)
1 pointsby 22 minutes ago
—discuss
Open Challenge List (openchallengelist.com ↗)
2 pointsby 24 minutes ago
1comments
Ask HN: So is AI taking our jobs?
1 pointsby 24 minutes ago
1comments
Flawed benchmarks (epoch.ai benchmark registry filtered by flawed)(epoch.ai ↗)
1 pointsby 26 minutes ago
—discuss
Anthropic builds wet lab for AI-driven drug discovery (qz.com ↗)
2 pointsby 26 minutes ago
2comments
Show HN: I compiled FreeRDP into a Flutter app so my phone can do RDP and SSH (xconnectapp.com ↗)
1 pointsby 27 minutes ago
—discuss
WeirdML v3 (htihle.github.io ↗)
1 pointsby 29 minutes ago
—discuss
Show HN: SIM Screener adds device frames to screenshots and recordings (simscreener.com ↗)
1 pointsby 30 minutes ago
—discuss
Show HN: Semantic-skill-kit:agent skills adapted from Google Modern Web Guidance (github.com/ulrischa ↗)
2 pointsby 33 minutes ago
—discuss

Grok Voice Transcribe 2.0

9 pointsby 49m agox.ai

Open original View on HN

47m agoHN ↗

No weights, sadly. Bit of a nothingburger release.

10m agoHN ↗

Word-level timestamps. Each word has precise start/end times and confidence scores.

Neat. This is missing from many models.

6m agoHN ↗

I don't understand how these benchmarks get things so wrong. Nova-3 is by far the best model on the market in WER. Azure's STT is atrocious. And yet every benchmark rates Azure's STT higher than Nova-3's.

Open thread on HN ↗