Hyperscalers are investing an unprecedented $600 billion in US capex this year, a tenfold increase from a decade ago, to meet insatiable demand for AI compute, with major players like Microsoft currently throttling services due to capacity constraints.
The dominant software business model is shifting from predictable, seat-based SaaS subscriptions to more volatile but potentially hyper-growth consumption-based or outcome-based pricing, challenging traditional valuation metrics.
AI is driving staggering productivity gains, exemplified by a developer achieving a 100x increase in output, which is leveling the playing field between incumbents and startups and forcing a re-evaluation of software company growth potential.
Venture capital firms are adapting to the AI wave by shifting investment focus towards the 'physical layer' (e.g., semiconductors, photonics) and grappling with how to underwrite and value companies experiencing unprecedented growth rates.
12 quotes
Concerns Raised
The sustainability of massive hyperscaler capex and whether capital markets will continue to support it long-term.
The shift from predictable SaaS revenue to volatile consumption-based models creates challenges for valuation and financial forecasting.
The rapid pace of AI development could make current technologies and investment theses obsolete in a matter of months.
Traditional software companies may see their growth rates impacted as AI levels the competitive playing field.
Opportunities Identified
Investing in the 'physical layer' of AI (semiconductors, photonics, networking) to support the massive infrastructure buildout.
AI-native companies are demonstrating growth rates far exceeding those seen in the SaaS era.
Massive productivity gains from AI tools can be leveraged across portfolio companies to accelerate growth and improve margins.
Companies providing infrastructure for the new consumption-based economy, such as billing and metering platforms.