Grok 4.5 Price War: SpaceXAI Undercuts Claude Opus by 75% to Dominate Coding Agents
SpaceXAI's Grok 4.5 launches at $2/$6 per million tokens, a 75% discount versus Claude Opus, igniting a Grok 4.5 price war that reshapes coding agent economics.
SpaceXAI's Grok 4.5 launches at $2/$6 per million tokens, a 75% discount versus Claude Opus, igniting a Grok 4.5 price war that reshapes coding agent economics.
Ukraine's combat humanoid robots gain state funding, but battlefield data from the Phantom MK-1 shows wheeled UGVs still dominate at scale in mud and EW conditions.
Nvidia's Vera Rubin rack cost hits $7.8 million as memory consumes a quarter of the bill. The Kyber delay to 2028 opens a window for competitors.
North Carolina joins Virginia in imposing a data center electricity tax on AI facilities, reflecting a broader backlash against subsidized AI infrastructure power costs.
Anthropic signed a $19B, 20-year TeraWulf Anthropic data center lease at a Kentucky campus, validating the bitcoin miner's pivot to AI infrastructure landlord.
Microsoft MAI models are replacing OpenAI and Anthropic in Excel and Outlook Copilot. The cost-driven shift raises quality concerns for enterprise subscribers paying the same price.
Apple OpenAI trade secrets lawsuit reveals coordinated theft of hardware designs as both companies race to build AI-powered devices. Strategic analysis inside.
The Microsoft Chevron power deal supplies 20-year behind-the-meter electricity to an AI data center, bypassing grid limits for large-scale compute.
Three EU regulatory actions in one week create urgency for enterprise AI teams as the AI Act compliance deadline of August 2 approaches with fines and pre-market rules.
OpenAI national security principles prohibit autonomous weapons and mass surveillance while deepening Pentagon and allied cyber defense partnerships.
Accenture Federal Services wins $821M task order for Pentagon War Data Platform integration, beating four bidders to serve as core integration partner for the CDAO-managed program.
Meta plans third-quarter production of its custom Iris AI chip, doubling compute to 14 GW by 2027 as the Meta Iris AI chip reduces reliance on Nvidia GPUs.
The AI data center backlash blocked $130 billion in projects during Q1 2026, forcing Amazon, Google, and Meta to rethink infrastructure strategy amid surging community opposition.
Perplexity adopts Nvidia Vera CPU for AI agent workloads, joining OpenAI and Anthropic in a strategic shift challenging Intel and AMD's server dominance.
OpenAI is sunsetting the ChatGPT Atlas browser less than a year after launch, pivoting to ChatGPT Work and its desktop app for enterprise agents.