$40 BILLION AI Company JUST went INSOLVENT | The AI Bubble.

The AI Bubble: A Deep Dive into Coreweave, Nvidia, and Potential Market Correction

Key Concepts:

• AI Capex Cycle: The investment cycle driven by Artificial Intelligence, encompassing infrastructure buildout (chips, data centers) and eventual application layer development.
• High Bandwidth Memory (HBM): A critical component in AI chip performance, currently experiencing supply constraints and rising costs.
• Coreweave: An AI data center provider facing potential financial difficulties, highlighted by revised debt agreements.
• Reuben Chips (Nvidia): Next-generation AI chips promising increased efficiency and potentially disrupting the current infrastructure cycle.
• Contagion: The risk of financial instability spreading from one company (Coreweave) to others (Oracle, Nvidia) within the AI ecosystem.
• Debt Service Coverage Ratio (DSCR): A financial metric used by lenders to assess a borrower’s ability to repay debt.
• Equity Cure: A provision in loan agreements allowing borrowers to inject equity to meet financial obligations.

I. The Looming Threat of an AI Bubble & Coreweave’s Financial Strain

The video opens with a discussion of widespread concerns about an “AI bubble” and its potential to burst, referencing media headlines and a significant $25 trillion debt bubble. The central focus quickly shifts to Coreweave, an AI data center company, and its recently released 8K filing revealing potential insolvency. The core argument is that the AI boom is fueled by a circular flow of money – investors funding companies like Coreweave, who then purchase chips from Nvidia, creating a potentially unsustainable cycle.

The analysis centers on Coreweave’s financial situation, specifically its debt obligations and revised credit agreements. The company has reduced its minimum cash requirement to $100 million and postponed testing of its Debt Service Coverage Ratio (DSCR) and contract realization ratio until late 2027. This is interpreted as a sign of distress, with banks essentially delaying scrutiny of Coreweave’s financials and relying on potential bailouts from companies like Nvidia. The filing explicitly allows for “unlimited equity cures,” meaning Nvidia could inject capital to cover debt obligations, effectively shielding Coreweave from covenant violations.

II. Coreweave’s Cycle & The Role of Nvidia

The video details the “Coreweave cycle”: investor money flows in, funding the purchase of AI chips (Nvidia, AMD), construction of data centers, and ultimately, hopefully, revenue generation. However, this cycle is hampered by construction delays and a significant amount of assets currently “in progress” (approximately 30%, or $6.9 billion).

A key indicator of Coreweave’s risk is the rising credit default swap (CDS) price for Oracle, seen as a safer proxy for Coreweave. The CDS price spiked to levels not seen since 2008, indicating increased investor concern about potential default.

Nvidia’s role is presented as both a potential savior and a potential disruptor. Nvidia has a $6.3 billion agreement to purchase unused data center capacity from Coreweave, offering immediate financial relief. However, the introduction of more efficient Reuben chips threatens to render existing infrastructure (and Coreweave’s investments in older chips) obsolete.

III. High Bandwidth Memory (HBM) & The AI Capex Cycle

The discussion expands to the critical role of High Bandwidth Memory (HBM) in AI chip performance. HBM is currently in high demand and experiencing supply constraints, driving up costs. This is a key component of the AI Capex cycle. The video explains the cycle’s stages: GPT moment (initial excitement), infrastructure explosion (chip purchases), and the eventual application layer.

The current stage is identified as late in the LLM (Large Language Model) infrastructure buildout phase, with HBM costs peaking. The video argues that rising HBM costs reduce margins for AI companies, potentially exacerbating financial pressures. The speaker positions the current situation as the seventh inning of the LLM cycle, but still early in the application era.

IV. Elon Musk, XAI, and Funding Dynamics

The video then pivots to Elon Musk’s recent funding round for XAI, initially denied as $15 billion but ultimately confirmed at $20 billion with Nvidia as a strategic investor. This is presented as a pattern – Musk initially downplaying funding needs, followed by a larger-than-expected raise.

The $20 billion raise is framed as sufficient to cover XAI’s projected $13 billion in burn rate for the current year. The speaker highlights the importance of strategic investors like Nvidia, who can anchor valuations and attract further investment.

V. Nvidia’s Position & The Future of AI

Nvidia is positioned as a likely survivor of a potential AI market correction. The company is not only benefiting from the infrastructure buildout but is also investing in the application layer, particularly through its self-driving technology (Nvidia FSD) and the acquisition/licensing of Grock, a language processing unit company.

The speaker believes Nvidia’s Reuben chips, promising a 10x reduction in token costs, will further solidify its dominance. While beneficial for Nvidia, this could negatively impact companies like Coreweave, which have invested in older, less efficient hardware.

The video concludes with a bullish outlook on the long-term potential of AI, emphasizing the importance of the application layer and the development of productive uses for AI technology.

VI. Personal Bias & Investment Strategy

The speaker discloses a bias towards the application layer of AI, citing a recent $10 million fundraising success for their own startup focused on real estate SaaS. They highlight the potential for AI to identify undervalued properties, creating opportunities for “net worth sniping.”

Notable Quotes:

• “I think AI completely saved the stock market.”
• “The banks are basically saying, ‘Hey, you know what? Um, we’re just going to go monkey no evil until Halloween of 2027.’” – Describing the banks’ willingness to delay financial scrutiny of Coreweave.
• “Daddy Jensen [Hong from Nvidia] comes around and gives you a bailout.” – Referring to Nvidia’s potential to rescue Coreweave.
• “Nvidia will be a survivor when this crash comes.”

Data & Statistics:

• Coreweave Debt: $7.5 billion total debt, $2.9 billion repayments, $4.75 billion net borrowings.
• Coreweave Cash Flow: Burning $500 million per month, $4.75 billion loss over 9 months.
• Coreweave Construction: 35% of assets are currently “in progress.”
• Nvidia Agreement: $6.3 billion agreement to purchase unused data center capacity from Coreweave.
• XAI Funding: $20 billion Series E funding round.
• XAI Burn Rate: $13 billion projected burn rate for the current year.
• Oracle CDS Price: Spiked to levels not seen since 2008, currently in the low 140s.

Conclusion:

The video presents a cautionary tale about the potential risks within the AI boom, focusing on the precarious financial situation of Coreweave and the broader implications for the AI infrastructure cycle. While acknowledging the long-term potential of AI, the speaker emphasizes the importance of understanding the underlying financial dynamics and the potential for a market correction. The introduction of more efficient technology from Nvidia, while beneficial for the company itself, could exacerbate the challenges faced by companies heavily invested in older infrastructure. The overall takeaway is that the AI cycle is maturing, and a shift towards profitable applications will be crucial for long-term sustainability.