Enterprise demand for sovereign AI favors Palantir's transformation model. In a two-pronged AI economy, businesses can retain their IP and use low-cost open models while frontier labs compete for the last increment of intelligence. x.com 

Palantir's reliance on pay-as-you-go service with refunds makes customer cancellations and sudden earnings declines or write-downs a serious risk as AI and high SaaS prices pressure the business. Its 10-Q warns of these risks. Anthropic can and will easily replace Palantir with an all-in-one platform under Democratic control of the legislative branches. x.com  x.com 

ServiceNow still looks too expensive even after falling 60-70% from its highs, given slowing growth, compressing margins, too much dilution and significant future AI risks. @realroseceline notes guidance drove the drop: the Armis acquisition brings temporary margin pressure, current remaining performance obligations are decelerating from 22% to 18%, and stock-based compensation has reaccelerated. x.com 

Palantir's ontology can make AI work verifiable beyond coding, helping enterprises adopt it more broadly. Its map of an enterprise's objects, relationships and rules could make even an average Nemotron model state of the art for that enterprise's own work. If results are good enough, firms will use them for sovereign AI; enterprises will not want to blindly run closed frontier models on codified context and give their know-how away. x.com 

Palantir's product thesis gains as Alex Karp signals the end of the GPT API era for enterprises, shifting the bottleneck from compute to deployment, which is exactly what Palantir sells and what the market hasn't priced. Per CNBC, Karp says OpenAI and Anthropic's token model is completely wrong as enterprise CEOs balk at the costs. x.com 

ServiceNow is a bad play for anyone trying to make a directional bet on AI. The better expression is AI infrastructure, where power, cooling, custom silicon, memory, lithography, silicon photonics, InP tools and connectivity silicon are each separate bottlenecks that are structurally short on supply. x.com 

Palantir gains as enterprises look for model-agnostic platforms rather than over-relying on any one AI lab, a lesson from the Anthropic/Mythos drama. It can be the neutral third party with deep AI expertise, mediating between AI labs and government, AI labs and the private sector, and government and the private sector. The $128 level has held well, which keeps the risk defined. x.com  x.com 

Palantir remains attractive, but CEO Alex Karp's sales pitch is the wrong strategy. At AIPCon he said customers will learn that model companies do not care about their business, only token use. The better message is that models keep getting more valuable, which makes Palantir's orchestration and proper deployment inside an organization even more valuable. x.com  x.com 

Palantir holds one of the two layers of AI that matter, structuring data into intelligence a model can interpret, while raw data holders like Salesforce and ServiceNow are the commodity. Replying to @StockMarketNerd, who argues models are commoditizing and value will shift to firms holding customer data: small benchmark gains feel shocking in real use, so models will not commoditize. Data does not matter, intelligence does. Bargain hunting is a losing game; you pay up for top talent. x.com 

Palantir's high-margin consulting layer is being commoditized as frontier models make complex data orchestration cheap and fast. Enterprises need not spend $10 million and 18 months on a rigid Foundry ontology when frontier models can parse messy data and build workflows in hours; Alex Karp's token-fee complaints reflect that. Its national-security moat is cracking as military operators use tools like Claude Code directly, and talent is fleeing to frontier labs. Palantir is a short. x.com