Devin Baker

I think of this as switchboarding , trying to get the right information to the right person. Someone’s got an empty seat/someone needs a ride. You’re getting into the history of plumbing/I know exactly the book you should read. Your cousin is moving to San Diego and doesn’t know anyone/my former rugby teammate lives there, maybe they can be... See more

We are building the most powerful tools we have ever had. Our temptation is a very American one: to engineer immediately. To ask of every new capability: What can I build with this? How do I deploy it? What can I do with it? How can I change creation with it? These are all necessary questions, but they are second questions. They are Genesis 2:15... See more

Webster, although he was a deeply religious man himself, reversed the order. He tended before he named. He decided what American English should be, and he built it. The result was efficient and clean and much faster. But this is what happens when you engineer before you discover. You get a useful product, but you lose the life and the weight of the... See more

We are entering an era where the vast majority of language produced will not have come from or been originated by human hands. These are often seen as words without lives — fluid, grammatically impeccable words that are efficient. But they are words that were never spoken, often about nothing, and for transient reasons. The model does not attend.... See more

New software companies emerged, too. Built on AI-native architectures with radically lower cost structures, they offered capabilities that didn’t exist in the prior generation. Enterprise software spending didn’t shrink for long; it got reallocated. By Q3 2027, total enterprise tech spending had recovered to 2025 levels but the composition was... See more

Enterprise software spending in the United States was roughly $500 billion annually in 2025. It was bloated, overpriced, and riddled with shelfware. Companies were spending millions on tools that 30% of their employees actually used. The AI-driven renegotiation and consolidation cycle cut that bill significantly. Where the bears lost the thread was... See more

The real estate example was instructive. The bears flagged the compression of buy-side commissions from 2.5-3% to under 1% as evidence of economic destruction. But the median home price in 2027 was $420,000. A 2-point commission reduction saved the buyer $8,400. Multiply that across 5 million annual home transactions, and that’s $42 billion per... See more

Consumer agents began handling purchasing decisions, subscription management, price comparison, and service negotiations. The bears modeled this as the destruction of the intermediation layer. They were right about the mechanism. Where they were wrong was in thinking that trillions of dollars in eliminated friction would simply vanish from the... See more

The bears saw this and modeled economic destruction. They were right about the mechanism — software margins compressed, enterprise spending was renegotiated, and the long tail of SaaS got repriced. Where they went wrong was in confusing the repricing of one sector with the collapse of the economy .