How great tech leaders identify and scale transformative technology

Executive overview

Most new technologies look useless until a single moment when you can personally experience their value. Identifying which technologies will reach that moment — before they do — is the core skill of a great tech leader and investor.

Mike Schroepfer, former CTO of Meta, distills this into three questions: How much headroom does the technology have to scale? Are there external tailwinds improving it without your effort? Does it solve a problem people actually care about?

Bet on technologies far from their theoretical limit, with external tailwinds, that solve a real problem people will pay to fix.

The three questions for evaluating breakthrough technology

1. The light speed test — how far is the current version from its theoretical maximum? A technology at 0.001% of its ceiling has far more room to grow than one at 99%.
2. Tailwinds — are external forces improving the technology without your direct work? In AI, compute got ~2x cheaper every 18 months via Nvidia and the chip ecosystem, independent of any AI team's effort.
3. Customer demand — does it solve a problem people genuinely care about? 3D TVs were a real technological advance; they failed because nobody wanted glasses on their couch.

Why AI looked obvious in hindsight but wasn't

• The ImageNet challenge was the signal: a neural net entered and outperformed every other approach by a margin that had never been seen before.
• The right question was not "is this impressive?" but "is this at the beginning or end of its runway?"
• The inputs powering AI — model size, dataset size, compute — could all be scaled thousands of times without any new invention.
• Early AI demos were terrible. The chatbot moment (ChatGPT) was when most people finally experienced the value themselves.
• Most people doubt a technology until they can touch it. That gap between "working" and "obviously useful" is where the opportunity lives.

Tackling scale: lessons from building Facebook's infrastructure

• In 2008, Facebook couldn't rent data center space — the financial crisis had halted construction.
• The solution was to build their own, hiring people who had the relevant background first.
• Core principle: work against the human instinct to do tractable tasks instead of hard ones.
• Identify the highest technical risks and work on those first, not the problems that feel comfortable.
• There is no getting away from hard problems — the only move is to run at them.

Why clean energy is the upstream problem for everything

• AI progress, clean water, air conditioning, manufacturing at scale — all require massive increases in energy.
• The world knows how to solve most resource problems; it does not know how to scale those solutions to 8 billion people without terawatts of additional clean energy.
• Solving sustainability means re-engineering tens of trillions of dollars of the global economy — too large for government or philanthropy alone.
• Startups are the mechanism: businesses investing in better, cheaper, cleaner products.

What GigaScale looks for in founders

• Relentless determination — building a company means a continuous stream of nos from investors, recruits, and customers. Getting 30 nos before one yes is normal.
• Rapid domain learning — the CEO job changes every day. The best founders combine humility ("I don't know this") with curiosity ("I'll figure it out fast").
• Neither trait can be reliably screened via resume — evaluation requires multiple meetings and deep reference calls.
• Bob Mumgaard (Commonwealth Fusion): plasma physicist, first-time founder, learned to operate a 1,000-person company from scratch.
• Matt Rogers (Mill): second-time founder; success at Nest didn't lead to complacency — he built a great team around himself again.

The Mill example: climate solutions that consumers actually want

• Mill is a kitchen bin that dries and grinds food waste into odourless coffee-ground-like material, emptied monthly.
• The consumer pitch is purely selfish: empty your trash less, and it stinks less.
• The climate impact is significant: food waste in landfill is a major source of methane and near-term warming.
• It generates real revenue — proof that environmental benefit and commercial success are not in conflict.