Forest Katsch retweeted
Here is SpaceX's new @Starlink Mini 2 with quick-swap battery. You can get up to 4.5 hours of use on a single charge.
This means you can get internet in the middle of nowhere and not have to plug anything in.
SpaceX has just officially unveiled its next-generation @Starlink Mini 2, which now comes with a built-in battery for the first time.
• Up to 4.5 hours of connectivity on a single charge with the quick-swap battery (or plug in USB-C).
• Weight: 3.53 Ibs (1.6 kg) including battery and kickstand
• Dimensions: 10.35 × 12.36 × 0.91 inches (263 × 314 × 23.1 mm)
• Field of view: 55°
• Wi-Fi: Wi-Fi 6
• Environmental rating: IP67
• Wi-Fi coverage: Up to 1,800 sq ft (167 m2)
• Battery: Removable, rechargeable 71 Wh battery
• Battery weight: 1.08 Ibs (490g)
• Battery attachment: Quick-release mechanism
• Charging: USB-C, 60W
• Battery power output: USB-C, up to 60W
• Power button: Built-in physical power button
• Battery indicators: LED status lights
• Kickstand: Integrated, foldable kickstand
• Tripod mount: Standard 1/4"-20 camera thread
• Charger: Included 60W USB-C wall charger
• Cable: Included 3-meter (10-foot) USB-C to USB-C cable
Starlink is now taking preorders in the U.S.: starlink.com/mini-2
Forest Katsch retweeted
this is bait and i’m getting baited but oh my god if people want to read your writing so bad they’re downloading and distributing it you should be overjoyed. that’s winning, that’s what winning looks like. you are being passed around as samizdat! you are the revolution! the only real objection i can imagine to this is if your name as an author is being stripped but otherwise this is free fucking publicity, people can look you up once AO3 comes back up, what the actual fuck
Forest Katsch retweeted
Knew shit was a wrap when the fucking SpaceX launch stream telemetry graphics came on screen 😭
Forest Katsch retweeted
Sad story time.
One of my friends got into a motorcycle wreck 1 year ago... just got out of the hospital. Hes quadriplegic now with minimal hand use. Used to be an amazing designer... and hes got fucking grit
He called me today and wants to build an exo to walk again. Between me and a few of yall I think we could get him going in 1 to 2 years, honest timeline.
I can buy the materials, I cant clone myself yet. Long horizon project, one that could improve the lives of millions.
Who's in? Whose willing to help build a mech that can carry a 150lb body in it?
Forest Katsch retweeted
the amount of compute left on the table by inefficient programming is shocking
Forest Katsch retweeted
Announcing Grok 4.5, our first model trained specifically for coding and agents. It was trained with Cursor and offers frontier intelligence at leading speeds and cost efficiency.
x.ai/news/grok-4-5
Forest Katsch retweeted
The xAI team working on Grok CLI has been absolutely awesome to work with. They were the ones who initiated this collab and they’ve driven it every step of the way.
Increasingly bullish on xAI and the future of Grok
Forest Katsch retweeted
It must feel horrendous to watch the thing you worked so hard on turn to dust. Sending love to all the Blue Origin folks going through that now 💙
Grok Voice is #1!
Announcing agentic performance benchmarking for Speech to Speech models on Artificial Analysis. We use 𝜏-Voice to measure tool calling and customer interaction voice agent capabilities in realistic customer service scenarios
Even the strongest Speech to Speech (S2S) models today resolve only about half of realistic customer service scenarios end-to-end - a meaningful gap relative to frontier text-based agents on the same tasks. Voice channels introduce significant complexity: challenging accents, background noise, and packet loss, all while requiring fast responses, consistency across long multi-turn conversations, and reliable tool use. Performance also varies considerably by audio condition: in clean audio some models perform notably better, but realistic conditions continue to pose a challenge. Conversation duration also varies meaningfully across models, with implications for both customer experience and operational cost.
About 𝜏-Voice:
Our Agentic Performance benchmark is based on 𝜏-Voice (Ray, Dhandhania, Barres & Narasimhan, 2026), which extends 𝜏²-bench into the voice modality to evaluate S2S models on realistic customer service tasks. It measures multi-turn instruction following, support of a simulated customer through a complete interaction, and tool use against simulated customer service systems. The simulated user combines an LLM-driven decision model with realistic audio synthesis: diverse accents, background noise, and packet loss modelled on real network conditions.
This complements our Big Bench Audio benchmark measuring intelligence and Conversational Dynamics (Full Duplex Bench subset) benchmark measuring conversational naturalness. Scores are the average of three independent pass@1 trials. We evaluate under realistic audio conditions using the 𝜏²-bench base task split across three domains:
➤ Airline (50 scenarios): e.g., changing a flight, rebooking under policy constraints
➤ Retail (114 scenarios): e.g., disputing a charge, processing a return
➤ Telecom (114 scenarios): e.g., resolving a billing issue, troubleshooting a service problem
Task success is determined by deterministic checks against expected actions and final database state, consistent with the 𝜏²-bench evaluator.
Key results:
xAI's Grok Voice Think Fast 1.0 is the clear leader at 52.1%, averaging 5.6 minutes per conversation, the second-longest overall. OpenAI's GPT-Realtime-2 (High) (39.8%, 3.0 min) and GPT-Realtime-1.5 (38.8%, 4.8 min) follow, with Gemini 3.1 Flash Live Preview - High close behind at 37.7% (3.8 min).
Speech to Speech is a fast evolving modality and we expect movement in rankings as we continue to add new models with these capabilities, and model robustness improves.
Congratulations @xAI @elonmusk! See below for further detail ⬇️

































