AI BriefingAI is slowing down
Key point
Whether astronomical revenue can be achieved to justify AI infrastructure investment, along with the uncertainty of measuring ROI, is emerging as a key risk.
Details
Generative AI infrastructure is projected to need to generate over $2 trillion in annual revenue by 2030 to justify data center investments and compute commitments. Building the planned 190GW of data centers could cost $9.5 trillion to $15 trillion, requiring $500 billion to $1 trillion in annual debt issuance, according to the analysis.
OpenAI is expected to burn through at least $852 billion by the end of 2030, while Anthropic would need to achieve $174 billion in annual revenue by 2029 to cover its compute costs. Currently, 89% of AI startup revenue is concentrated in these two companies, making additional demand essential to sustain the scale of compute being built out.
As companies shift to token-based billing, they are struggling to measure ROI (return on investment) for AI spending. In response, Uber, T-Mobile, Brex, and others have begun implementing cost controls such as setting per-employee token spending limits.
If major customers like OpenAI fail to cover their enormous compute costs, there is also a risk that companies like Oracle, which are building out large-scale infrastructure, could face funding difficulties.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Anthropic Pacing to $100B ARR; Rumors of OpenAI Solving Navier-Stokes Spark ControversyOpenAI and Anthropic's RSI Strategy and the Uncertainty of AI Profitability