This week on Upside, Dan Bowyer and Mads Jensen of SuperSeed and Lomax Ward of Outsized Ventures unpack a moment where AI infrastructure, enterprise adoption and market risk are all moving at once.

Nvidia is laying out a path toward a $1 trillion AI market, driven by major advances in inference performance.

At the same time, hyperscalers are investing at unprecedented levels — with AI capex increasingly supported by debt rather than free cash flow. But the real shift is happening higher up the stack.

The AI race is moving away from pure model performance and toward distribution, enterprise control and monetisation.

This episode explores:

• Nvidia’s roadmap and the scaling of AI infrastructure
• Hyperscaler capex and the return of balance sheet risk
• Why the AI battleground is shifting to enterprise
• OpenAI’s monetisation challenge and strategic positioning
• The growing gap between AI capability and adoption
• Where value actually accrues in the AI stack
• And how hyperscalers are reshaping startup opportunities

This isn’t just another AI cycle.

It’s infrastructure, capital and business models being rewritten at the same time.

Chapters

00:00 Intro02:00 Nvidia GTC and inference leap07:00 The trillion-dollar AI question12:00 Hyperscaler capex and leverage18:00 Models vs distribution23:00 OpenAI’s strategy28:00 Enterprise AI battleground34:00 Market risk and concentration39:00 Capability vs adoption45:00 Where value accrues50:00 Startups vs hyperscalers55:00 Europe policy signals

Upside is a weekly deep dive into the forces shaping European venture, AI, defence and deep tech.

Hosted by:Dan Bowyer (SuperSeed)Mads Jensen (SuperSeed)Lomax Ward (Outsized Ventures)

About Upside

Nvidia’s $1T AI Bet, Hyperscaler Risk & The Real AI Battleground | Upside

EUVC