The biggest AI companies on earth just spent $6.5 billion on a confession. In the last 60 days: - OpenAI: raised $4B+ for a new deployment company - Anthropic: launched a $1.5B services firm with Blackstone and Goldman Sachs - AWS: put $1B of its own cash into a forward deployed engineering org - Microsoft: answered two days later with $2.5B and 6,000 engineers - Google Cloud and Salesforce: hiring the same roles as fast as they can Whats the role they're all hiring for? It's the FDE. And that brings up a ton of criticism. I keep hearing: "both CAN'T be true. Either AI is easy to use and makes everyone productive, or you need embedded engineers because it's hard. Pick one." But, both ARE true and they are answers to 2 different problems. Individual productivity is easy now. A person with a chat window gets immediate gains, no engineering needed. Enterprise productivity is a completely different animal. The same capability has to run across 10,000 employees, touch customer data, pass security review, integrate with a 20-year-old ERP, and produce numbers a CFO will sign. That's not a prompting problem... It's a systems problem. We've seen this before. Nobody argued "spreadsheets are easy, so we don't need ERP systems." Individual tools and enterprise systems have always been different businesses. The gap between them is where 95% of AI pilots die (MIT's number, not mine). The labs know it. None of that $6.5 billion buys better models. All of it buys deployment. If you're a leader: the easy wins are the floor. Budget for the systems work, not just the licenses. The chat window was the demo. Enterprise systems are the business. (That gap is what we work on at @tenex_labs . It's real, and it's wide.)