Wall Street is freaking out over OpenAI now...

Summary

The video argues that OpenAI’s rapid growth is being financed by unusually large losses and infrastructure commitments, while its customers, suppliers, investors, and potentially the public become increasingly connected to its financial outcome. Its central business question is whether revenue can eventually scale enough to justify the cost of running and expanding AI services.

Business model and financial performance

• ChatGPT reportedly had more than 900 million weekly users, with the latest model costing around $20 per month for paying users. The video presents this as exceptional consumer-product adoption, but says usage also drives substantial computing and energy costs.
• For 2025, the video cites:
  • Revenue rising from $3.7 billion to $13.07 billion—roughly tripling.
  • Total costs and expenses of $34 billion and an operating loss of $20.92 billion.
  • About $2.60 in spending for every $1 of revenue.
  • A reported net loss of $38.53 billion, some of which the video says reflects accounting rather than cash outflow.
  • $19.18 billion in R&D spending and $5.73 billion in sales and marketing, with sales and marketing up more than fivefold.
• The video contrasts OpenAI’s losses with Amazon’s largest annual net loss of $1.4 billion, arguing that OpenAI’s operating losses are on a different scale.
• A Deutsche Bank analysis cited in the video projected $143 billion in cumulative losses by 2029. It says OpenAI had about $50 billion in assets, nearly half in cash.
• The business model has a cost-scaling challenge: unlike conventional software, each additional AI query consumes compute, electricity, cooling, and chip capacity. More users can bring more revenue, but also increase operating costs.

Strategy, governance, and financing

• OpenAI was founded in 2015 with a nonprofit mission. The video says the expense of developing frontier AI pushed it toward commercial financing, beginning with a shift in 2019, and that it became a for-profit company in October 2025.
• The video recounts the 2023 leadership crisis: Sam Altman was briefly removed as CEO, employees threatened to leave, and Altman was reinstated. It characterizes the outcome as reducing constraints on the company’s ability to raise and spend capital.
• OpenAI’s high valuation is portrayed as a way to finance further investment, not just as a measure of current performance:
  • A SoftBank-led funding round valued it at $300 billion and was associated with a further $40 billion infrastructure investment.
  • An employee share sale in October 2025 reportedly raised $6.6 billion from private investors at a $500 billion valuation. The video says about 600 employees sold shares, averaging roughly $11 million each.
  • At $500 billion, the video estimates the valuation at about 25 times projected annual revenue, compared with a stated 5–15 times range for many established software companies.
• The video frames this valuation strategy as a bet on future AI dominance: investor capital enables more compute, which is intended to produce stronger models and future revenue.

Infrastructure and operating commitments

The video presents OpenAI’s spending commitments as a major execution risk:

• In September 2025, it reportedly committed to at least 10 gigawatts of AI data-center capacity using Nvidia systems.
• Nvidia CEO Jensen Huang is cited as estimating that one gigawatt of AI data-center capacity can cost up to $60 billion to build.
• By late 2025, OpenAI reportedly had around $1.4 trillion in data-center and computing commitments—about $70 of long-term commitments for every $1 of revenue.
• Internal projections cited in the video anticipated about $856 billion in compute and infrastructure spending by the end of 2030.
• The video also reports that OpenAI agreed to spend $300 billion on Oracle cloud services and says it paid Microsoft $17.2 billion for research, infrastructure, and other costs, while Microsoft paid OpenAI $303 million in return.

A key operating challenge is that infrastructure must be financed before it can support future revenue. The company therefore depends on external capital, supplier financing, and continued investor confidence.

Financing playbook and risk concentration

The video describes a circular financing loop in the AI supply chain:

1. Nvidia invests in OpenAI.
2. OpenAI uses funding to buy cloud and infrastructure services, including from Oracle.
3. Oracle buys Nvidia chips to provide that capacity.
4. The original funding flows back toward Nvidia as chip sales, while the companies record growth in different forms.

Additional details cited in the video include:

• Nvidia reportedly planned to invest up to $100 billion in OpenAI in stages tied to data-center capacity coming online.
• Analysts had reportedly identified more than $800 billion in similar circular arrangements across the AI supply chain by 2026.
• Nvidia was also reportedly in talks to guarantee roughly $250 billion in financing for a 10-gigawatt campus in Ohio. An SEC filing is cited regarding Nvidia guarantees for land, power, and buildings at part of the site. The video says the chips for the site could add as much as $350 billion in costs.

The business risk is counterparty concentration: if OpenAI cannot pay, suppliers and lenders could face losses, and Nvidia’s growth could become tied to the solvency of major customers. The video compares this structure with equipment vendors’ lending to internet startups before the dot-com bust, while noting that Nvidia is much larger and financially stronger than those historical examples.

Potential effects beyond the company

The video identifies several channels through which AI infrastructure risks could affect others:

• Corporate earnings and investors: Microsoft reportedly said OpenAI-related investment losses reduced net income by $3.1 billion in one quarter, or $0.41 per share. The video also argues that paper gains from private AI valuations can affect reported earnings and market sentiment.
• Electricity and infrastructure: PJM, which serves about 65 million customers across 13 states and Washington, D.C., reportedly projects summer demand to rise about 30 gigawatts by 2030 compared with 2025, much of it driven by data centers. Capacity-auction prices rose from $28.92 per megawatt-day for 2024–25 to $329.17 for 2026–27. PJM’s independent market monitor attributed 63% of the increase to data centers; the NRDC estimated consumers could pay an additional $163 billion through 2033.
• Public policy: The video recounts OpenAI CFO Sarah Friar’s comments about government guarantees for AI-chip financing, followed by her clarification that OpenAI was not seeking a government backstop. It also says OpenAI proposed extending chip-manufacturing tax credits to data centers and sought grants, cost-sharing, loans, or loan guarantees.
• Government ownership: In July 2026, Sam Altman reportedly proposed a 5% U.S. government stake, valued in the video at about $42.6 billion, as a way to let the public share in AI’s upside. The video argues that such ownership could also expose taxpayers to losses.
• Public-market access: The video says OpenAI confidentially submitted a draft S-1 to the SEC in June 2026, while Altman said it would not go public that year. It discusses a possible 2027 IPO and Nasdaq’s fast-entry rule, under which sufficiently large new listings could enter the Nasdaq-100 within 15 trading days. More than $30 trillion tracks major indexes, potentially creating automatic demand from index funds and pensions if OpenAI qualified.

Frameworks, processes, and actionable takeaways

• Scale first, monetize later: The video compares OpenAI’s approach with the Amazon playbook, but argues that the size of the losses makes the comparison uncertain. The practical test is whether future revenue and gross margins can outgrow infrastructure and operating costs.
• Infrastructure-to-revenue discipline: Track not only user growth and sales, but also the cost of serving each query, capacity utilization, energy costs, and the capital required for each increment of capacity.
• Financing and counterparty-risk review: Map who funds each project, who receives the spending, and who bears losses if demand or repayment falls short. The circular Nvidia–OpenAI–Oracle example illustrates why recorded growth across several companies may depend on the same underlying funding.
• Separate committed spending from realized spending: Distinguish announced commitments, guarantees, and projected costs from cash already deployed and revenue already earned.
• Stress-test the growth thesis: The video’s key scenario is whether OpenAI can reach the cited $350 billion in annual revenue by 2030. If it does, current losses could look like a successful long-term investment; if it does not, suppliers, lenders, investors, and infrastructure customers may absorb the consequences.

The video does not present a formal operating framework such as OKRs or a go-to-market plan. Its main business lesson is to evaluate AI companies through unit economics, capital efficiency, infrastructure returns, and the resilience of their financing chain, not user growth or valuation alone.

Presenter and sources mentioned

Presenter: Josh Risser, narrator for The Infographics Show.

Sources and people cited: Deutsche Bank; an unnamed analyst; Jensen Huang and Nvidia; Microsoft’s quarterly filing; Oracle; SoftBank; the Wall Street Journal; the SEC; Sarah Friar; David Sacks; Sam Altman; Elizabeth Warren; Bernie Sanders; PJM and its independent market monitor; the Natural Resources Defense Council; Nasdaq; S&P 500; FTSE Russell; and Anthropic.