⚡TΞSLASΞNTINΞL🔋

If there is momentum behind brining TikTok.US into and under the umbrella of Musk entities (however a deal could be struck) with this configuration, it would be complicit with US information security guidelines. *** Background *** ByteDance Ltd., based in Beijing incorporated in the Cayman Islands, parent company of TikTok: Ownership: 60% institutional investors, 20% employees, 20% founders with Zhang Yiming controlling via dual-class shares. TikTok Inc., California subsidiary, manages U.S. operations. TikTok Ltd., Cayman entity, oversees global operations. Beijing Douyin Information Services Co., Ltd. manages Douyin, the Chinese counterpart of TikTok. Discussions for selling TikTok's U.S. operations: ByteDance retaining 80% (not for sale) Oracle & Walmart potentially owning 20% (stalled) this is up for potential sale / controlling interest - Can this 80 : 20 ratio be titled toward the US based controlling company's favor ? Important Core Point: The U.S. business would include operations and user data but *** not the core algorithm *** -> how can the US Based Controlling entity gain access to the algorithm Connection to China: Douyin vs. TikTok: Separate content moderation due to legal differences. *** Data and Algorithm*** ByteDance in China controls the algorithm; U.S. data privacy is a concern and potential risk to National Security. ( Can the US TikTok control backend and data back flows to the Douyin controlled backend? ) Operational Link: ByteDance's tech support to TikTok suggests deep integration, complicating divestiture as the algorithm isn't easily separable. ==========================> Because of the structure of the current terms and the portions which would be under control of either a Musk entity or other US controlling entity, not sure the US should allow TikTok to operate in the US (Unless there are very strong changes to how the deal would be structured, and enable the US Controlling interests to thoroughly dissect the backend engineering of the systems that house and US data and those that operate the algorithm. If the US controlling interest can seize control of the algorithm, and thorough control of the back end engineering, this might be a dangerous national security risk, if not thoroughly vetted. Hopefully some deal could be struck to bring the US Tiktok under a musk entity control, because we know Elon will see to it that some of the security concerns above and others not even privy to everyone would be properly vetted. Either way, US must tread carefully metadata on US citizens and their corresponding cross-relational connection to what would be China based military interests needs to be carefully monitored.

Full thrust 33-engine static fire with Super Heavy V3

Gemma 4: Now up to 3x Faster. ⚡ Same quality, way more speed. Our new MTP drafters allow Gemma 4 to predict multiple tokens at once, effectively tripling your output speed without compromising intelligence.

Tesla Semi

Cybercab is now testing across the U.S. Tesla’s driverless platform is expanding beyond a single city, with sightings and test activity reported in multiple states. This points to a broader validation phase, not a limited pilot. The goal is clear. Stress test the system in different traffic patterns, road designs and real-world edge cases before scaling. If this pace continues, the rollout could move faster than many expect.

This Chinese developer launched Llama 70B locally on a MacBook on a plane and for a full 11 hours without internet ran client projects. He was sitting by the window on a transatlantic flight with a MacBook Pro M4 with 64 GB of memory. WiFi on board cost $25 for the flight. He declined. No cloud API, no connection to Anthropic or OpenAI servers, no internet at all. Just a local Llama 3.3 70B on bf16 and his own orchestrator script. The model runs through llama.cpp. Generation speed, 71 tokens per second. Context around 60,000 tokens. Memory usage, 48.6 GiB out of 64. Battery at takeoff, 3 hours 21 minutes. And he gave the orchestrator this system prompt before takeoff: "You are an offline orchestrator running on a single MacBook. There is no network. The only resources you have are local files in /Users/dev/work, the Llama 70B inference server at localhost:8080, and a battery budget of 3 hours 21 minutes. Process the queue at /Users/dev/work/queue.jsonl (one client task per line). For each task: draft → run local evals → save artefact to /Users/dev/work/done/. Save context checkpoints every 12 tasks so you can resume after a battery swap. Stop only on empty queue or when battery drops below 5%." So the system knows exactly what resources it is running on. It knows it has no connection to the outside world for the next 11 hours. It knows it has finite memory and a finite battery. It knows the human will not intervene until the plane lands. The system runs in 1 loop. Takes a task from the queue, runs it through inference, saves the artifact, writes a checkpoint. Task after task, just like that. And only when the battery drops below 5% does the orchestrator automatically pause, waits for the laptop to switch to the backup power bank, and continues from the last checkpoint. Here is what the system actually writes in his log during the flight: "saved context checkpoint 8 of 12 (pos_min = 488, pos_max = 50118, size = 62.813 MiB)" "restored context checkpoint (pos_min = 488, pos_max = 50118)" "prompt processing progress: n_tokens = 50 / 60 818" "task 37016 done | tps = 71 s tokens text → /Users/dev/work/done/proposal_westside.md" Outside the window, clouds, blue sky, and no WiFi. On the tray, 1 MacBook, an open terminal on 2 screens, and an inference server on localhost. From what I have observed, this is the cleanest offline AI workflow I have seen in the past year: 11 hours of flight, $0 for WiFi, and the entire client queue closed before landing.

AN ENTIRE AI COMPANY JUST GOT OPEN-SOURCED • 147 agents across engineering, marketing, sales, finance, and more • Runs on Claude, Cursor, Copilot, Gemini — deploy with one command • Started on Reddit, now 88K stars, 14K forks, fully MIT licensed

Falcon Heavy landing

The Most Comprehensive Codex tutorial on the internet right now even OpenAI's own team said it's one of the best in under 2 hours you'll learn how to use Codex + GPT-5.5 (imo the strongest combo available right now) from skill and plugin setup, to automated workflow building, to multi-threading practical demos if you're looking for the best practices for vibe coding, this is the one to watch big shoutout and respect to @rileybrown

I have Claude Code fully integrated into my smart glasses Everywhere I go I am now vibe coding When you are talking to me in real life, I am not paying attention to you. I am nodding, but I have no idea what you are saying. That's because I am silently shipping. I am watching the green lines of the terminal scroll by at all times. I am silently giving commands to my glasses on what to build next. I am reaching record levels of LOC shipped. There are multiple lakes in Michigan that no longer have water in it because of me I am using up my Claude plan faster than anyone else on earth. Nobody can match my velocity. If you are building a competing product, it is over. Give up. Unless you have Claude Code being projected into your cornea at all hours of the day, you simply have no shot. GGs.

SpaceX has released a new mini-documentary about Starship and version 3 of the rocket. Incredibly well produced!

Falcon Heavy

CATL, the world’s largest battery manufacturer, has unveiled its 3rd-generation Shenxing LFP battery. • 10% to 35% charge in 1 minute • 10% to 80% in 3 minutes & 44 seconds • 10% to 90% in 6.5 minutes. • Even at temps as low as -30°C (-20°F), CATL said it can recharge from 20% to 98% in as little as 9 minutes • CATL claims the new LFP battery can retain over 90% of its original health after 1,000 ultra-fast charging cycles • World-record internal resistance of 0.25 milliohms, ~50% lower than the average • New self-heating pulse technology for better charging performance in cold weather • New Cell Shoulder Cooling Technology, which is 20% more efficient “We always deliver what we promise,” Gao Huan, the company’s CTO, told reporters today at an event in Beijing.

The creator of Claude Code teaches more about vibe-coding in 30 minutes than most tutorials do in hours. Save this — it'll change how you build forever.

Falcon flies every few days. You can see launches in person from Florida or California.

It’s a start. Vastly more will be needed to extend consciousness to the stars.

🎀 Yeah, I have been bothering you with more than 100,000 posts 🥳🥂🎊🎉🪇🙌🎀 Can't tell you how grateful I am to have you all. ❤️

Congrats to the @Tesla_AI chip design team on taping out AI5! AI6, Dojo3 & other exciting chips in work.

That’s my goal