--- title: The More You Buy, The More You Lose | daily.dev description: A detailed financial and critical analysis of the AI infrastructure spending cycle, arguing it is fundamentally unsustainable. Hyperscalers (Google, Meta,... canonical: https://daily.dev/posts/the-more-you-buy-the-more-you-lose-kgzqyw1ts twitter:card: summary_large_image twitter:site: @dailydotdev og:type: website og:site_name: daily.dev og:title: The More You Buy, The More You Lose | daily.dev og:description: A detailed financial and critical analysis of the AI infrastructure spending cycle, arguing it is fundamentally unsustainable. Hyperscalers (Google, Meta,... og:url: https://daily.dev/posts/the-more-you-buy-the-more-you-lose-kgzqyw1ts og:image: https://api.daily.dev/og/posts/kgZqYw1TS.png og:image:alt: The More You Buy, The More You Lose og:image:width: 1200 og:image:height: 630 og:locale: en --- > ## Documentation Index > Fetch the complete documentation index at: https://daily.dev/llms.txt > Use this file to discover all available pages before exploring further. # The More You Buy, The More You Lose **[Where's Your Ed At](https://daily.dev/sources/wheresyoured)** · 40 min read · 1 upvotes · 0 comments ## Summary A detailed financial and critical analysis of the AI infrastructure spending cycle, arguing it is fundamentally unsustainable. Hyperscalers (Google, Meta, Amazon, Microsoft) have collectively spent over $1 trillion in capex on AI with no disclosed AI revenues, while taking on hundreds of billions in on- and off-balance-sheet debt. NVIDIA's Blackwell GPU efficiency claims have been walked back from 25x to 10x, yet no AI company has become profitable. The piece dissects a proposed $250B NVIDIA-OpenAI-SoftBank circular financing deal as emblematic of the broader problem: money flowing between interconnected investors with no real external demand. CoreWeave's junk-level bond spreads, Google's century bond losing 10% of value, and Oracle's credit downgrade are cited as market signals of distress. The author argues that AI's actual utility is vastly overstated, that corporate AI adoption is driven by executive signaling rather than productivity, and that the entire ecosystem is a self-reinforcing bubble that will eventually collapse when capex spending slows. ## Full article daily.dev links to this article rather than hosting it. Read it at the original source: ## Questions this post answers ### What is the NVIDIA and OpenAI $250 billion financing deal with SoftBank's SB Energy about? NVIDIA was reportedly in talks to guarantee $250 billion in financing for a 10GW data center allegedly being built by SoftBank affiliate SB Energy, covering lease payments and construction costs, while separately discussing financing $350 billion in GPUs for the same facility. SB Energy would use the OpenAI lease contract to raise debt and eventually take the company public, benefiting both SoftBank and OpenAI's equity positions. _Anyone evaluating AI infrastructure deals can follow circular financing arrangements like this one on daily.dev._ ### How much debt have hyperscalers taken on for AI data centers and how risky is it seen as being? Meta, Google, Amazon, and Microsoft added over $850 billion in property, plant, and equipment and over $307 billion in on-balance-sheet debt since March 2022, reaching $557 billion total, plus roughly $1.35 trillion in off-balance-sheet debt via SPVs and undelivered contracts, according to Nikkei. Bond markets are pricing this poorly, with CoreWeave's bonds trading at a 756bps spread versus 540bps at issuance, and Google's 100-year bond already down 10% in value. _Developers and engineers tracking AI infrastructure risk can follow debt and capex trends like these on daily.dev._ ### Why did NVIDIA's Blackwell GPU cost savings claims get revised downward? NVIDIA CEO Jensen Huang originally claimed Blackwell GPUs would reduce LLM inference operating cost and energy by up to 25x, but that figure was later pared back to 10x in blog posts citing case studies with private inference providers that don't disclose margins and aren't profitable, representing roughly a 150% overstatement of the original claim. _Engineers evaluating GPU generation claims for inference cost can track this kind of scrutiny on daily.dev._ ## Similar posts on daily.dev - [Premium: The Hater's Guide To AI Debt \(Part 1\)](https://daily.dev/posts/premium-the-hater-s-guide-to-ai-debt-part-1--p9dmod0fg) · Where's Your Ed At · 3 upvotes · 0 comments - [What must happen for AI’s trillion-dollar gamble to pay off](https://daily.dev/posts/what-must-happen-for-ai-s-trillion-dollar-gamble-to-pay-off-w9wk3lfgo) · MIT Technology Review · 0 upvotes · 0 comments - [AI Is Too Expensive](https://daily.dev/posts/ai-is-too-expensive-snj1hxice) · Where's Your Ed At · 86 upvotes · 22 comments --- Tags: [#ai](https://daily.dev/tags/ai), [#llm](https://daily.dev/tags/llm), [#nvidia](https://daily.dev/tags/nvidia) [View this post on daily.dev](https://daily.dev/posts/the-more-you-buy-the-more-you-lose-kgzqyw1ts) ```json {"@context":"https://schema.org","@graph":[{"@type":"Organization","@id":"https://daily.dev/#organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180},"sameAs":["https://twitter.com/dailydotdev","https://github.com/dailydotdev","https://www.linkedin.com/company/daily-dev-ltd"]},{"@type":"WebSite","@id":"https://daily.dev/#website","url":"https://daily.dev","name":"daily.dev","publisher":{"@id":"https://daily.dev/#organization"},"potentialAction":{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https://daily.dev/search?q={search_term_string}"},"query-input":"required name=search_term_string"}}]} {"@context":"https://schema.org","@type":"TechArticle","headline":"The More You Buy, The More You Lose","url":"https://daily.dev/posts/the-more-you-buy-the-more-you-lose-kgzqyw1ts","mainEntityOfPage":{"@type":"WebPage","@id":"https://daily.dev/posts/the-more-you-buy-the-more-you-lose-kgzqyw1ts"},"datePublished":"2026-07-28T16:32:20.030Z","dateModified":"2026-09-14T06:16:05.159Z","description":"A detailed financial and critical analysis of the AI infrastructure spending cycle, arguing it is fundamentally unsustainable. Hyperscalers (Google, Meta,...","image":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/574adc906abb37956ea31d3307bf03c0?_a=AQAEuop","thumbnailUrl":"https://media.daily.dev/image/upload/f_auto,q_auto/v1/posts/574adc906abb37956ea31d3307bf03c0?_a=AQAEuop","isAccessibleForFree":true,"articleSection":"Where's Your Ed At","inLanguage":"en","publisher":{"@type":"Organization","name":"daily.dev","url":"https://daily.dev","logo":{"@type":"ImageObject","url":"https://daily.dev/apple-touch-icon.png","width":180,"height":180}},"author":{"@type":"Organization","name":"Where's Your Ed At","logo":"https://media.daily.dev/image/upload/s--_fiBlTSf--/f_auto/v1737645971/logos/wheresyoured","url":"https://daily.dev/sources/wheresyoured"},"commentCount":0,"discussionUrl":"https://daily.dev/posts/the-more-you-buy-the-more-you-lose-kgzqyw1ts","interactionStatistic":[{"@type":"InteractionCounter","interactionType":{"@type":"LikeAction"},"userInteractionCount":1},{"@type":"InteractionCounter","interactionType":{"@type":"CommentAction"},"userInteractionCount":0}],"keywords":"ai,llm,nvidia","timeRequired":"PT40M"} {"@context":"https://schema.org","@type":"BreadcrumbList","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https://daily.dev"},{"@type":"ListItem","position":2,"name":"Where's Your Ed At","item":"https://daily.dev/sources/wheresyoured"},{"@type":"ListItem","position":3,"name":"The More You Buy, The More You Lose"}]} {"@context":"https://schema.org","@type":"FAQPage","@id":"https://daily.dev/posts/the-more-you-buy-the-more-you-lose-kgzqyw1ts#faq","mainEntity":[{"@type":"Question","name":"What is the NVIDIA and OpenAI $250 billion financing deal with SoftBank's SB Energy about?","acceptedAnswer":{"@type":"Answer","text":"NVIDIA was reportedly in talks to guarantee $250 billion in financing for a 10GW data center allegedly being built by SoftBank affiliate SB Energy, covering lease payments and construction costs, while separately discussing financing $350 billion in GPUs for the same facility. SB Energy would use the OpenAI lease contract to raise debt and eventually take the company public, benefiting both SoftBank and OpenAI's equity positions. Anyone evaluating AI infrastructure deals can follow circular financing arrangements like this one on daily.dev."}},{"@type":"Question","name":"How much debt have hyperscalers taken on for AI data centers and how risky is it seen as being?","acceptedAnswer":{"@type":"Answer","text":"Meta, Google, Amazon, and Microsoft added over $850 billion in property, plant, and equipment and over $307 billion in on-balance-sheet debt since March 2022, reaching $557 billion total, plus roughly $1.35 trillion in off-balance-sheet debt via SPVs and undelivered contracts, according to Nikkei. Bond markets are pricing this poorly, with CoreWeave's bonds trading at a 756bps spread versus 540bps at issuance, and Google's 100-year bond already down 10% in value. Developers and engineers tracking AI infrastructure risk can follow debt and capex trends like these on daily.dev."}},{"@type":"Question","name":"Why did NVIDIA's Blackwell GPU cost savings claims get revised downward?","acceptedAnswer":{"@type":"Answer","text":"NVIDIA CEO Jensen Huang originally claimed Blackwell GPUs would reduce LLM inference operating cost and energy by up to 25x, but that figure was later pared back to 10x in blog posts citing case studies with private inference providers that don't disclose margins and aren't profitable, representing roughly a 150% overstatement of the original claim. Engineers evaluating GPU generation claims for inference cost can track this kind of scrutiny on daily.dev."}}]} ```