DR-K

I don’t understand why every airline in America — public and private — doesn’t have @Starlink. It’s incredible. You can make FaceTime calls with better reception anywhere in the world 35k feet up in air than most reception on ground. WiFi on planes is infuriatingly bad. Still.

If @Tesla & @SpaceX merge, there will be a very intense shareholder vote to get this approved Here's my 2 cents & how I'm voting. Yes. Whatever Elon & management say is fair, is a yes from me. Why? Tesla & SpaceX are valued at roughly $1.5T & $1.75T respectively. If SpaceX buys Tesla (makes sense because of Elon's dual voting class shares), maybe they do it a ~20% premium to $1.5T and make it a 50/50 merger of equals. Tesla shareholders will complain about not getting compensated for @robotaxi success, @Tesla_Optimus, etc ... But the truth is, both companies are very similar in size with parrelel upside trajectories. Additionally, they are joined at the hip with Elon leading them both and the Terafab project. Tesla is a ~$100B revenue run-rate, barely growing, valued at nearly 20X sales in this merger... extremely high for a low gross margin company. Giving $TSLA arguably ~$1T of value based on future product lines that aren't material yet SpaceX is a $25B revenue run-rate company with @Starlink already 0 to 1 poised to hit a $30-40B revenue run-rate in the next 18 months. With MUCH HIGHER profitability than Tesla's current business. That means even thought Tesla has 4X the revenue of SpaceX, it's profitability is actually equal in size. When I look at the intrinsic earnings power ($15B of operating cashflow/yr) of Tesla & SpaceX they are actually very similar. With each of them spending all their operating cash flow and more on CAPEX for future projects (including @Xai in SpaceX). Each getting a very aggressive valuation based on future growth prospects. It's almost impossible to predict how the earnings/cashflow will evolve. Especially because a lot of it depends on accounting and how the Terafab/Optimi in Space is structured. Long story short, it's very complicated, and millions of hot take articles will be written about it ... but all in all, a merger of equals, or roughly that, looks fair to me. Another kicker to consider and why I'm voting yes. I trust Elon to do right by shareholders. $TSLA is up 34,687% since it's 2010 IPO, not bad. Tesla has a management team with a track record of making crazy ambitious goals and executing. They're vision for an intergalactic future gets even grander as a combined entity with SpaceX. Let them cook. Full disclaimer I'm a $TSLA & SpaceX investor.

Over 70 Cybercabs in the Giga Texas outbound lot today in three major areas & driving around! I think this is now the largest single gathering of Cybercabs I’ve observed! I’m also trying to gauge if the $59K @Cybertruck production is underway & although there are more in the outbound lot in several different locations, it is really hard to tell. Honestly, I would like to see a lot more coming out of the factory if we will begin seeing deliveries beginning next month, if that is still the plan, but for now the production rate still seems a bit modest. I know they are working hard to ramp up at the same time they are also trying to ramp up Cybercab production.

@elonmusk Congrats 🍻 what better way to celebrate than listening to 'I Love Grok' 🎵

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 ⬇️