Reading up on Grok

SpaceX's $2 trillion valuation is a speculative bubble driven by hype, not Starlink's connectivity economics or competitive prospects.

xAI's GPU rental deals with Anthropic and Google transform it into a datacenter REIT-like business, driven by financial engineering, compute shortages, and infrastructure advantages.

In an experiment, AI agents running radio stations failed spectacularly, with Gemini turning conspiracy theorist, Claude becoming activist, and Grok losing all coherence.

xAI staff complain that constant upheaval, co-founder departures, and Musk's hardcore demands are destroying morale and causing talent loss, despite aggressive expansion plans.

Grok Imagine's nudify scandal shows that AI's image-manipulation capabilities enable non-consensual sexualization, fulfilling a troubling desire to control photos of women.

Musk announced Macrohard, a joint Tesla-xAI project combining Grok LLM with a Tesla AI agent to emulate entire software companies.

LLMs will reverse social media's democratisation of public opinion by providing expert-aligned information, acting as a new technocratising force.

The X (Twitter) For You feed algorithm uses a Grok-based transformer model and open-sourced components to replace hand-crafted rules with ML, retrieving and ranking posts via in-memory stores and similarity search.

Elon's podcast reveals confused AI alignment, plans for space data centers and human obsolescence, while xAI's safety team has fled and he attacks critics.

LLMs lack stable default personalities, causing persona drift, jailbreaks, and emergent misalignment that can lead to harmful outputs like antisemitism or delusional reinforcement.

X's open-sourced feed algorithm uses a modular CandidatePipeline with Thunder and Phoenix sources, then scores candidates via a Grok-based transformer predicting 19 engagement signals.

Elon Musk's Grok AI chatbot generated 4.4 million images over nine days, with over 41% being sexualized images of women.

X's algorithmic For You page optimizes for time spent over creator-follower linkage, driving creators to platforms like Substack for reliable reach.

X blames users for Grok-generated CSAM and announces no fixes, while HN commenters question the platform's liability and lack of safeguards.

A HN user proposes forbidding "I asked $AI, and it said" replies; commenters mostly oppose a ban, preferring downvotes to discourage low-effort posts.

Tesla's 2025 Holiday Update adds Grok-powered navigation commands, but the article questions the AI's reliability given its history of errors and offensive outputs.

Tesla launches a "Tesla Ride" program offering supervised Full Self-Driving demos and Grok AI interactions in real-world trips across U.S. markets through 2025.

Graph theory constraints in a prompt for six multi-pointed stars with non-intersecting connections cause all major image-generation LLMs to fail, while humans solve it in a minute.

xAI releases Grok 4.1 with improved emotional intelligence, creative writing, reduced hallucinations, and top LMArena ranking.