Saturday, November 14, 2026

SpaceX Starship Flight 15

 Starship Flight 15 launch, ship catch & CNN's first reactions - YouTube


In an unprecedented maneuver, the Starship spacecraft came sweeping down to a pinpoint landing within the arms of SpaceX’s Mechazilla launch tower. 

This marks the first time in the history of rocketry that the upper stage of a rocket has returned home from orbit to a safe, upright landing — potentially paving the way for it to be used again. 

SpaceX attempted to guide the Super Heavy rocket booster back to a pinpoint landing at its South Texas launch tower, called Mechazilla, but the maneuver was ultimately unsuccessful. The company redirected the booster to splash down in the Gulf.


AI: Grok vs Opus 5.5; C vs JAX, SGLang

surprising they didn't use Rust :)

 Elon Just Called a Rival's AI a "Masterpiece" - YouTube @ Brighter with Herbert - YouTube

Competitive Status: Elon Musk recently admitted that xAI's Grok model is currently behind leading competitors, singling out Anthropic's Opus 5.5 as a "masterpiece." However, he expressed confidence that the upcoming Grok 5 could catch up or surpass it within 2 to 3 months.

Architecture Overhaul: xAI completely abandoned JAX and SGLang, rewriting its entire reinforcement learning and inference stack from scratch in C/C++ (with a minimal amount of C++).

Engineering Roots: Musk noted that the new AI training and inference stack draws directly from engineering solutions developed for Starlink, referencing low-level practices like writing custom headers and manually calculating time (including accounting for leap seconds).




Fight for AI control?

The "sudden" fall of Nvidia CEO (and founder) and the biggest AI "boom" salesman
appears to be more than personal misconduct issue, but a proxy for fight for control.

Will AI tech be "centralized" or "decentralized"?

There are a lot of competing interests, and a lot to lose and possibly gain, 
so politics and big business are making and using the turmoil to establish or maintain control.

Complicated and consequential... Can't be just ignored, even if most people can't do much about it...

And some AI-slop fun :)

《Wallstreet Boys - I Want It That Way》 - YouTube


Watters: This is turning into the country’s biggest political scandal - YouTube

Jensen Huang Scandal Explained Like You're 5 - YouTube

DOJ Admits "No Spy" — Right After Nvidia's $1.7T Crash - YouTube



After coup against Nvidia CEO, US admits Huang lover is NOT Chinese spy - YouTube


1. Jensen Huang Ousting & Allegations (1:20 Opens in a new window )

  • The Sequence of Events: The discussion details how two major news stories broke in rapid succession, culminating in Jensen Huang being removed as CEO of Nvidia after just one day of intense public and regulatory fallout.

  • The Allegations & Memo: The segment covers leaked classified documents and a whistleblower report regarding how Nvidia chips were tracked or moved in relation to export restrictions and international markets.

  • Board Action: The panel analyzes the board's decision-making process, focusing on how executive leadership and the board of directors had to evaluate their fiduciary duties under heavy political and regulatory pressure.

2. Market Impact & The AI Economy (31:22 Opens in a new window )

  • "The Bank of Nvidia": The podcast breaks down Nvidia's critical financing and hardware role as the core engine funding the broader enterprise AI capital expenditure and infrastructure buildout.

  • Geopolitical Restrictions & Strategy: Discussion centers on global export controls, how chips are deployed, and the operational friction international hardware restrictions introduce into global supply chains.

3. Additional Timestamps & Topics Discussed in the Episode






  • Jensen Huang's Departure & NVIDIA's Future: The panel dives into the shockwaves caused by Jensen Huang leaving his role, examining how NVIDIA plans to handle its succession and what this means for the company's dominance, market strategy, and leadership moving forward.

  • Geopolitics and Chip Restrictions: Discussion centers around the intersection of AI hardware, national security, and policy—highlighting tensions between industry leaders, intelligence agencies, and regulatory frameworks regarding international markets (such as China).

  • AI Safety vs. Acceleration ("Doomers" and Guardrails): The panel tackles the ongoing debate over safety guardrails versus the hyperspeed development of AI, with Emad Mostaque and Peter Diamandis debating whether bureaucratic or regulatory frameworks can realistically keep pace at scale.

  • Recursive Self-Improvement (RSI): The conversation touches on advanced AI milestones, touching upon perspectives—such as those from Demis Hassabis—regarding how recursive self-improvement is shaping the approach toward advanced intelligence.