Anthropic cozies up to Elon Musk, IBM makes its case to win in AI, and Cerebras has big IPO plans
Anthropic’s desperate handshake with Elon Musk exposes the raw cash hunger driving Big AI, even as IBM quietly builds the enterprise path that actually works.
Anthropic just inked a deal to consume every last GPU in Elon Musk’s Colossus supercomputer. The same Musk who spent months calling the company “Misanthropic,” labeling its Claude model racist, and comparing its logo to an anal sphincter. Money, it turns out, is an extraordinary solvent for grudge.
Both sides need the arrangement. Anthropic is burning capital to keep up with its “ridiculous growth.” Musk’s xAI apparently cannot fill the data center with enough users of Grok, so SpaceX gets revenue that makes an IPO look less terrifying. The source itself calls it a “marriage of convenience” likely headed for the rocks. That cynicism feels earned.
While hyperscalers chase one another’s shadows, IBM used its Think conference to argue it finally possesses the operating model, hybrid infrastructure, and agent framework to move AI beyond prototypes into daily enterprise use. Arvind Krishna’s keynote laid out “AI-first enterprises, hybrid as default,” and quantum shifting “from science to engineering.” ServiceNow, meanwhile, positioned itself as the control tower for the coming wave of agents. These are the quieter, less meme-driven bets that large companies actually intend to run on.
The earnings scorecard this week told its own story. Datadog, Palantir, AMD, and Akamai beat and rose sharply. Cloudflare, CoreWeave, Arm, and Kyndryl missed or guided poorly and paid for it in the stock. Despite endless SaaS-pocalypse warnings, the winners were the firms delivering measurable infrastructure or platform value rather than vapor. Cerebras, actually profitable, plans to raise $3.5 billion in its IPO at a $27 billion valuation, a reminder that hardware reality still matters.
OpenAI and Anthropic are both rushing into joint ventures with private-equity firms to sell AI services to enterprises. That pivot signals they recognize consumer chat is not the endgame. Yet the funding frenzy continues elsewhere: Sierra at $15 billion, Moonshot AI at over $20 billion, ElevenLabs topping $500 million in annualized revenue. The capital is still flooding in even as the strategic illegibility argument gains traction, Brian Halligan’s warning that founders are commoditizing their own moats by making everything legible to AI.
The contradiction sits in plain sight. Hyperscale 3.0 is being drawn on whiteboards while the empires that already run corporate backbones, IBM, ServiceNow, SAP, methodically wire AI into the systems that cannot fail. The Musk-Anthropic spectacle makes better headlines. The real contest for who wins enterprise AI will be won in the less glamorous architecture layer. Watch which vendors still have customers willing to bet their operating model on them next quarter.
Original reporting: SiliconANGLE News.
More in Business
- IMF Chief Warns of Economic Divergence and Debt Risks Amid Artificial Intelligence Boom
Kristalina Georgieva suggests that while AI could significantly boost global growth, it simultaneously threatens to widen inequality and exacerbate inflationary pressures and public debt burdens.
- California Limits Utility Profits and Expands Home Energy Role
Gov. Gavin Newsom signed laws capping utility returns on lower-risk projects and accelerating virtual power plant development.
- Lyft to Pay $272.5 Million in Record California Wage-Theft Settlement
Largest such settlement in state history resolves claims that Lyft misclassified drivers as independent contractors from 2016 to 2020, denying them core wage protections.