The All-Hands That Rewrote the Narrative
On February 11, 2026, xAI published a 45-minute internal all-hands meeting video directly to X — not leaked, not reported second-hand, but deliberately broadcast to the world. The timing was surgical: Multiple sources had just reported on the Tuesday night gathering, priming a negative cycle around mass co-founder departures. Musk's response was to detonate a bigger story on top of it. The video revealed a company energizing a GPU cluster equivalent to 1 million H100s at its Memphis supercluster — drawing over one gigawatt of power, backed by the world's largest Tesla Megapack installation — and outlined plans for orbital data centers and a lunar manufacturing base using a mass driver, an electromagnetic rail system, to launch AI satellites into deep space.
This is the story behind the spectacle.
The Merger Is a Capital-Structure Fix, Not a Love Story
Ten days before the all-hands, on February 2, SpaceX announced its acquisition of xAI in a $1.25 trillion deal — the largest corporate merger in history. SpaceX is valued at $1 trillion; xAI at $250 billion. A planned 2026 IPO targets up to $1.5 trillion in valuation, with shares expected to price around $527 and up to $50 billion to be raised publicly.
The financial reality is blunt: xAI burns approximately $1 billion per month. SpaceX generates $15–16 billion in annual revenue and $8 billion in profit. Strip away the interplanetary rhetoric and this is a cash-burning frontier AI lab attaching itself to a profitable launch monopoly to survive long enough to matter. The space-compute narrative — orbital data centers targeting 100–200 gigawatts annually, lunar factories scaling theoretically to 1,000+ gigawatts — is real long-term intent, but it also functions as valuation architecture: it reframes the AI race away from pure model benchmarks, where xAI faces stiff competition from OpenAI and Google, toward vertical integration of energy and launch infrastructure where no rival can follow.
Sophisticated investors should treat the moon factory as a 10–15 year option with extreme variance, not an underwriting thesis. The synergy that actually moves IPO pricing is the triangle of X's distribution (1 billion installs, $1 billion in subscription ARR), SpaceX's financing credibility, and demonstrable AI product revenue — none of which requires a single rocket launch.
The Four Pillars Reveal What xAI Is Actually Selling
The reorganization — announced the same day as the video — restructured xAI into four divisions: Grok Model and Voice under Aman Madaan; Infrastructure under Manuel Kroiss; Applications and Products; and Research and Development. Read between the lines: this is a company optimizing for shipping and monetization ahead of an IPO, not research purity.
The four product pillars tell the same story. Grok Main targets an "everything app" with forecasting and voice. The Coding division pursues recursive self-improvement — Grok training Grok — aiming to bypass traditional compilers entirely by end of 2026. Imagine already generates 50 million videos and 6 billion images per month, claiming to surpass all competitors combined. Macrohard, the most ambitious and least underwriteable, targets full digital emulation of entire companies — automating end-to-end functions across software, law, and design. Treat Macrohard as a call option, not a base case; it is also the easiest place to hide missed timelines.
Standalone apps — X Chat (an encrypted WhatsApp rival to be open-sourced) and X Money (entering external beta, targeting all user financial transactions) — represent the fastest path to revenue proof. Payments, however, will be gated by regulators, making X Money's timeline a key variable to watch.
The Co-Founder Exodus Is the Real Red Flag
Six of xAI's original twelve co-founders have now departed, including Tony Wu and Jimmy Ba this week alone. Musk framed all exits as "push, not pull" on X, writing that the reorganization "unfortunately required parting ways with some people." The Financial Times reported Jimmy Ba's exit was driven by rising technical team tensions under pressure to accelerate model performance against OpenAI and Anthropic.
The bull case is org hardening — the SpaceX model, where execution-first culture produced 300-plus launches in 2025. The bear case is institutional collapse. The critical signal to monitor over the next two quarters is not founder departures, which are survivable, but whether churn migrates to mid-level engineering managers and staff engineers. That is where roadmaps die.
The IPO Is the Moment of Truth
Public markets will not price this on AGI promises alone. The combination will be sold on three pillars: SpaceX's cash engine as the multiple anchor, AI as a growth kicker requiring credible revenue signals, and platform bundling across Starlink, X, and Grok. If xAI's burn rate fails to show a clear downward slope against revenue growth, markets will re-rate the combined entity as SpaceX with a costly sidecar — compressing the multiple fast. Regulatory risk is a parallel threat: antitrust scrutiny of a $1.25 trillion vertical integration, export controls on space technology, and payments licensing for X Money all represent non-trivial tail risks that could derail the story before it reaches public investors.
The valuation surged 340x in 26 months — from $673 million at Series A in November 2023 to $230 billion at Series E in January 2026, drawing Nvidia, Fidelity, Andreessen Horowitz, Sequoia, BlackRock, and sovereign wealth from Qatar and the UAE. That trajectory is either the most impressive private-market run in history or the most elaborately constructed FOMO in the AI era. The IPO will settle the argument.
not investment advice