329
Audio & Video Production317
Software Development244
Automation & Workflow209
AI Infrastructure & MLOps149
Marketing & Growth188
Writing & Content Creation197
Data & Analytics120
Design & Creative149
Customer Support122
Photography & Imaging140
Voice & Speech132
Sales & Outreach111
Operations & Admin86
Education & Learning120
Investor Martin Casado says AI model building is less about rare talent and more about money, data, and computers, and he warns cheap funding may not last.
In short: Andreessen Horowitz investor Martin Casado says making AI models is getting less mysterious, but funding for the biggest AI labs may get harder.
Martin Casado, who leads an AI investment team at venture capital firm Andreessen Horowitz, said recent progress in artificial intelligence feels as big as the industrial revolution. He also compared AI’s impact to the early internet, because it changes how people communicate and how businesses work.
Casado argued that building AI models (the systems that learn patterns from lots of data) is not as difficult as many people think. In his view, the main ingredients are access to large amounts of data, powerful computers, and enough money to pay for both. He said he often sees teams without famous credentials build strong models when they can afford the resources.
He also warned that today’s biggest AI companies may not be able to raise “cheap money” forever. Casado said AI models can lose value quickly because new versions arrive fast, sometimes within three to six months. That means companies may need to spend heavily again and again just to stay current.
Casado thinks the AI industry could shift as investors become less willing to fund huge spending. If that happens, more of the business value may move “downstream” to smaller models and to apps that use AI for specific jobs, like customer support or coding (similar to how electricity created value in appliances, not just power plants). For everyday users, this could affect which AI products survive, how much they cost, and who controls the tools people use at work.
Source: Financial Times