In just the last 90 days: 1. Grok 4.3 — barely top 10. “xAI is dead beyond compute leases.” 2. Grok 4.5 — massive comeback. “Maybe a chance. But Grok will never catch up to the frontier.” 3. Grok 4.6 — they’re frontier. “Still not top 3.” 4. Grok 4.7 — now top 3 in frontier coding, while being much cheaper & faster. 5. Grok 4.8 next month..... The model machine is just starting up. Grok will be the workhorse of the upcoming agentic era.

Mwahaha, he's so miserable and evil that he wants to: 1. Give the world clean, abundant energy. 2. Bring AI & robotics abundance 3. Cure blindness & disabilities 4. Preserve freedom of speech. 5. Make humanity multi-planetary so our species survives & wins. Truly evil stuff. 😈

SpaceX just dropped $60B to acquire Cursor, diluting shareholders by ~3%. Was it worth it? Yes. Cursor was a stealth major AI lab ready to emerge & Elon took it first. Here are the 5 core reasons why: 1. Cursor is the #3# coding AI platform globally, behind only Anthropic & OpenAI: - 5M+ active coding users - 50,000+ enterprises & engineering teams - 64% of Fortune 500 (~70% of Fortune 1000) - 100M+ lines of enterprise code written daily 2. Core value: data Quality data is AI’s scarcest asset, and SpaceX AI lacked it, blunting its massive compute edge. Cursor sits at the center of millions of top engineers using the world’s leading models (Claude, GPT, Chinese models). It sees what they accept, reject, rewrite, and ship across real codebases. This creates a continuous feedback data loop of exactly what works. SpaceX AI absorbs that data to learn best behaviors across every major model and finally turns its compute advantage into frontier models that win. 3. Cursor team One of the most underrated elite teams in AI: they built the first at-scale IDE before Claude and shipped a third-ranked coding model that was 4-6x smaller, 10-60x cheaper, and nearly 3x faster in just 2 months. Results of the partnership so far: By combining Cursor data + massive compute, Grok 4.5 jumped 40% in capability overnight; Grok 4.6 added another 13%, reaching near parity with GPT-5.6 sol and Claude fable 5. Grok 4.7 is expected to hit #1# in weeks at an almost impossible catch-up speed. They also helped build Grok Build Claude Code/Codex competitor, all this in just three months. Now Post-merger, SpaceX AI can ship top models monthly by scaling compute against Cursor’s growing data trove at full throttle. 4. Revenue Cursor grew from $2B ARR (Feb) → $3B (late April) → $4B (early June)—~19% average monthly growth. At that rate, it compounds to ~$8.4B in 4 months and ~$12.3B in 6 months, putting it at $8–12B ARR by Q4. As the #3# coding tool worldwide, rising AI adoption accelerates usage further. With coding expected to be mostly automated by 2027 in a $1T+ annual industry, Cursor can sustain 20%+ monthly growth (or higher) and reach tens of billions in ARR as a top-3 player & market share holder. More than paying for its cost in 3 years. 5. Their own top model series: Post-merger, Cursor is assumed to have secured a guaranteed 100k GPUs for its own training. The team has emphasized they have not cancelled the Composer series and plan to release Composer 3 soon. Known for coding-focused models built on the back of Kimi, with Composer expected to be built from scratch. Composer 2.5 (last drop) was the 3rd-best coding model at 25% of competitor size, 10-60x cheaper, and much faster(ideal for the upcoming agentic era). With unlimited compute, they can finally build large models that seriously challenge the overall frontier.

Here is the full SpaceX Q2 $18.4B Ai mostly Capex Breakdown and full 2026 deployed compute breakdown: – Of the $18.4B total capex: $1.17B funded space operations (Starship, Starlink, production scale-up, launch infrastructure), while the majority, $15.8B, went into AI compute deployment. The $15.8B AI spend added 400 MW of nameplate capacity (1.0 GW → 1.4 GW), equivalent to ~220k GB300s (~130 kW per 72-GPU NVL72 rack) at $5.18M per deployed rack, a 25%+ discount to typical market deployment costs. – Total installed GPUs reached: 990k–1M (from ~770k+ in Q1: 220–230k Colossus 1 + 550k+ Colossus 2) note earlier Q1 filings did not fully reflect Colossus 1. – 2 GW+ year-end target per Elon: Reaching 2 GW of nameplate capacity requires 600 MW more deployed. Using Q2 averages ($3.95B per 100 MW for 55k GB300 GPUs), this implies $23.7B (23.7% of cash at hand) for 600MW in additional spend over the next two quarters which would add 330k GB300s & potentially including early Rubin systems, still at below market prices per deployed rack. – Projected year-end GPU count: 1.32M+ GPUs (using 2.3 GW): 220k+ in Colossus 1 550k+ in Colossus 2 550k in Macrohard's All of this totals ~3M+ H100-equivalents, making it the largest known AI compute deployment globally. – Compute Leasing: 450–465k GPUs already committed to Anthropic, Google, and Reflection AI, plus capacity for the new $20.8B+ unspecified cloud deals, estimated at 400k+ GPUs. Total to be leased: 860k+ GPUs, generating $48.6B ARR (with $27.8B confirmed annualized). – Remaining for SpaceXAI: 400k+ mostly GB300s (~1.2M H100-equivalents) for internal AI workloads and additional rentable capacity.