Editor's Note

The Federal Reserve raised interest rates by a quarter point in a unanimous vote, ending a three-year pause and choosing to fight inflation first. At the same time, Big Tech and frontier labs pushed forward with hundreds of billions in data center commitments and new consumer software. The conflict between five percent bond yields and capital-hungry AI roadmaps defined the entire month.

Inflation and Oil Force the Fed's Hand

Central bank accommodation officially reversed this month. The Federal Open Market Committee raised the target range for the federal funds rate by 25 basis points to 3.75% to 4.00% in a 12-0 decision.

Fed Chair Kevin Warsh explained the move bluntly, stating that "inflation is too high and has been for too long." Warsh noted that the 12-month change in total PCE prices likely reached around 3.6% in August. August headline CPI registered at 3.4% year over year, while August PPI rose 5.4% year over year, driven higher by energy.

Surging crude prices kept inflation pressures alive across the period:

• Brent crude pushed past $100 a barrel, briefly exceeding $105 to $108 following U.S. strikes on Iranian targets, Houthi attacks on Saudi infrastructure, and disruptions near the Strait of Hormuz.
• U.S. retail diesel hit record levels of $6.20 per gallon.
• European central bankers also tightened policy, with the European Central Bank raising rates by 25 basis points to 2.50%.
• The Bank of Japan lifted its policy rate to 1.25%, reaching a 31-year high.

The bond market reflected that pressure immediately. The U.S. 10-year Treasury yield climbed above 5%, touching 5.116% after strong services PMI data reached 58.7. Treasury Secretary Scott Bessent's $6bn government debt buyback disappointed traders who had anticipated $8bn to $10bn. Markets ended the period pricing roughly a 70% probability of another Fed rate hike in October, even as Donald Trump demanded rates of 1 percent or half a percent.

AI Infrastructure Spending Escalates Into Hundreds of Billions

While the cost of money jumped, the race to build computing capacity accelerated.

The financial scale required to sustain frontier AI models came into sharp focus:

• OpenAI forecasts cumulative negative free cash flow of $278 billion through 2030, alongside $856 billion in compute and infrastructure spending. OpenAI expects revenue to climb from $36 billion this year to $350 billion in 2030.
• Anthropic expects to exit 2026 with $100 billion in annual recurring revenue. It agreed to pay SpaceX approximately $1.25 billion per month through May 2029 for computing capacity at the Colossus 1 facility.
• Oracle posted a backlog of $664 billion in remaining performance obligations, booking over $30 billion in AI cloud contracts in Q1 while generating negative free cash flow of -$5.4 billion on capital expenditures of $28.5 billion.
• Broadcom reported Q3 AI semiconductor revenue of $16.7 billion, up 221% year over year, and guided Q4 AI revenue to $21.7 billion. Broadcom laid out targets of roughly $58 billion in FY26, $115 billion in FY27, and $230 billion in FY28 AI semiconductor revenue.
• Nvidia CEO Jensen Huang declared that the AI buildout is "not tens of billions, but tens of trillions," pointing to CoreWeave and Nebius as critical capacity providers.

Higher borrowing costs are starting to influence infrastructure financing. CoreWeave is paying 8.3% on roughly $35 billion of debt, pushing interest expense toward approximately $900 million per quarter. Nebius saw borrowing rates rise from 2.3% to 5.5% across three quarters. As central banks hold rates higher, the cost of capital is becoming as decisive as hardware demand.

Software and Silicon Shift Toward Consumer Monetization

To justify mounting infrastructure budgets, tech leaders moved to test actual commercial products.

Meta introduced Muse, its consumer personal AI assistant powered by the Muse Spark foundation models. The app reached 2.8 million downloads in its first 12 days and hit number one on Apple's U.S. App Store. Meta set pricing with a free tier and subscriptions at $20 or $100 per month.

To drive transactions, Meta announced integrations with Shopify via Shop Pay across all Shopify stores, alongside partnerships with PayPal, Instacart, and Expedia. Chief AI Officer Alexandr Wang confirmed Meta is exploring taking a cut of agent-related shopping transactions, though no concrete plan has been settled.

Hardware and chip architectures are adjusting to these workloads:

• Meta plans to begin deploying its custom MTIA 450 Arke chips in data centers in the first half of 2027, while MTIA 500 Astrid is expected to finish design work in roughly a month and enter data centers by the end of 2027.
• Bank of America estimates Meta could save roughly $8.5 billion in 2027 by deploying its own silicon.
• AMD crossed a $1 trillion market capitalization as investors anticipated that agentic AI would lift demand for inference hardware.
• Apple unveiled its first foldable smartphone, the iPhone Duo, starting at $1,999, powered by the A20 Pro chip using Micron LPDDR5X memory.

Counter-Thesis and Risk Watch

Contrarian commentary across the period focused on sovereign debt stress, AI financing risks, and commodity shocks:

• George Gammon argued that the market sits in an AI bubble, pointing to 10-year Treasury yields above 5%, 30-year mortgages above 7%, and rising default risks in commercial real estate and subprime credit.
• Shay Boloor noted that OpenAI's gap between $856 billion in projected spending and $278 billion in cash burn means the buildout depends entirely on capital markets remaining open and cheap for five more years.
• Wall Street Millennial pointed to circular financing risks around Nvidia data center leases and warned of off-balance-sheet commitments.
• Palantir CEO Alex Karp claimed on X that OpenAI will never go public because public markets cannot absorb frontier AI liabilities.
• Geopolitical Economy Report highlighted that U.S. Strategic Petroleum Reserve levels fell to 285 million barrels, their lowest level since November 1982, leaving little cushion against Middle East supply disruptions.

Looking Ahead

Can OpenAI and Anthropic secure the hundreds of billions in outside funding their multi-year roadmaps require with benchmark rates approaching 4%?

Will consumer applications like Meta's Muse generate meaningful software and transaction revenue before depreciation on current data center spending hits corporate earnings?