Topic

blog

A collection of 6 issues

The Incomplete Map: Why Materials Science Won't Have Its AlphaFold Moment

AlphaFold is a powerful interpolator, not an extrapolator. Savage's small vs large world framework explains why - and why the real prize in materials AI is building tools that know what they don't know.

Fail Fast, Learn Faster - Rethinking Talent Up-Skilling

I built a decentralized micro-fund to upskill 400+ postdocs. It funded great science, but data revealed broken incentives and AI vulnerabilities. Here is my post-mortem on failing fast in leadership, and redesigning a high-signal skills accelerator.

Marketing Discovery: Whither PhDs?

Peter Drucker said marketing makes selling superfluous. Today? We are effectively cold-calling geniuses to accept near-minimum wages for a 20% shot at a job. The legacy PhD model is broken. It’s time to stop selling an academic lottery ticket and start engineering a levered asset.

Scaling the 11th Century Start-up, or, Why Universities Break People

53% of academics show signs of depression. This is a role design failure, not a resilience issue: we are managing 21st-century experts with medieval guild structures. To fix the crisis, we must abandon the 'Lone Wolf' myth and re-engineer the role. Don't blame the players; change the game.

The £13B non-profit: Valuing the modern University

"University isn't worth it?" Applying a corporate valuation model to UK HE reveals a £13bn asset - more than Man Utd & City combined. It’s time to stop managing universities like charities and start valuing them as sovereign utilities.

Equity for Equations: The Rise of Venture-Backed Physics

I started 2026 with a grant rejection letter. Meanwhile, 'Neolabs' like Periodic and SSI are raising billions without a product. We are witnessing the birth of a new asset class: the venture-backed Institute for Advanced Study.