The AI Ecosystemby Mindy Wu Open the map →
About the AI Ecosystem

Who actually pays for AI?

You do, a few dollars at a time. Shopping, feeds, meals, trips, classes, paychecks, doctors and chatbots all run on AI. This map follows that money down through labs, clouds, chips, factories, power and rockets to a quartz mine in North Carolina.

Open the map → Zoomable, with every company and link.
0layers
0players
0links
0bottlenecks ◆

How to read it

1 · Start wideSix colour-coded zones, numbered in reading order, from what people spend to raw materials. Every company is one icon.
2 · Zoom inScroll, pinch or click a layer. The row opens to every company, then names and tickers.
3 · Click a companySee what it does and the opportunity. Amber: used by. Teal: depends on.
4 · Hover for helpDotted terms and layer names explain themselves. Pick an industry view (AI model, semiconductors, space, power, biotech), filter by headquarters country, or show only ◆ bottlenecks. Press / to search.

Views, countries and bottlenecks

An AI model, physically

Chips, memory, servers, power, buildings and minerals. Use the Highlight bar to light up each component.Open the AI model view →

Semiconductors

Designers, fabs, tools and materials down to raw quartz.Open the semiconductor view →

TSMC & Taiwan

About 40 Taiwan-based players: TSMC, design services, packaging and test, local tool makers, fab builders, substrates, servers and cooling, plus TSMC’s customers and foreign suppliers.Open the TSMC & Taiwan view →

Biotech

Drug makers, lab tools and contract factories, plus the chips and glass they depend on.Open the biotech view →

Where it is made

Each company shows its home country as a two-letter code. Filter to one country to see how concentrated a view is.Show Taiwan in semiconductors →

Guided journeys

Follow one everyday thing all the way down.

Four ways into this map

Own a piece

Public companies carry a ticker and quote link. Know which layer you understand, and whether you pay for a bottleneck or a story.Show public stocks →

Work inside it

Fabs, power, cooling and grid building are short of skilled people. Chip, ML and infrastructure roles span every layer.Show bottlenecks →

Build on it

Each layer has a build angle. Private companies are potential partners, customers or competitors.Show private companies →

Finance or govern it

Data centres run on private credit and sovereign money, and the bottom of the chain is shaped by export controls.Show other items →

Every layer here began as someone’s bold bet: a food company holds the key film inside AI chips; a rocket company rewrote launch costs. Be ambitious, be curious, and look for the layer others ignore.

Caveats

Educational, not investment advice. Facts are to about mid-2026 and some will have moved. Outlooks are qualitative; use each panel’s quote link for current figures.

Links are typical, publicly known relationships, simplified, not a list of contracts. Logos come from each company’s own site; trademarks belong to their owners.

Spot an error or a missing company? Tell me and I’ll add it.