Since Leaving Washington, Elon Musk Has Been All in on His A.I. Company (www.nytimes.com)

Elon Musk used a rare companywide meeting to lay out an intensified vision for xAI — the two‑year‑old start‑up behind the Grok chatbot — framing its mission as building “maximally truth‑seeking” AI and even previewing plans to take on Microsoft with a project he dubbed “Macrohard.” The 90‑minute presentation, given as Musk has shifted more of his time to xAI since stepping back from Washington, underscores his hands‑on role: long stints at the Palo Alto office, visits to xAI’s data center, and a push to align the company’s work with his other ventures (Tesla, SpaceX). xAI was most recently valued at about $120 billion, signaling strong investor interest in the bet. For the AI/ML community, Musk’s renewed focus is significant on several fronts. A public, mission‑level insistence on “truth” suggests xAI may prioritize model objectives, data curation and evaluation metrics that target factuality, robustness to adversarial misinfo, and verification — shifting engineering tradeoffs away from solely capability benchmarks. Rivalry with big tech raises competitive pressure on model scale, infrastructure and safety practices, while Musk’s dual leadership across companies and his rhetoric about government impotence highlight governance and alignment debates: who sets standards for truth, and how will xAI’s designs affect downstream use in transportation, space and broader commercial ecosystems?