Why Sequoia Thinks AI Services Will Crush Software in the $6 Trillion Battle

Sequoia just flipped the SaaS playbook on its head — they're betting services will eat software's lunch.

Here's the truth: we're witnessing the biggest shift in professional services since the internet.

The key insight? It all comes down to intelligence versus judgment.

Intelligence work follows rules. Writing code, processing claims, drafting NDAs, medical coding - complex rules, but still rules. AI has crossed the threshold where it can handle most intelligence work autonomously.

Judgment is different. It's knowing WHICH feature to build next. When to ship before it's ready. Whether that candidate will fit your culture. This requires experience, taste, and pattern recognition built over years.

Here's what's fascinating: software engineering got automated first because it's primarily intelligence work. That's why over half of all AI tool usage happens in coding, while other professions are still in single digits.

But every profession is moving along this spectrum.

The smart money isn't building better tools anymore. They're selling the outcome directly. Crosby doesn't sell legal software - they draft your NDA. WithCoverage doesn't sell insurance tools - they GET you insured.

The autopilot playbook is brilliant: start where outsourcing already exists. If companies already pay external providers for the work, you're just a better vendor. No reorg required.

Today's judgment becomes tomorrow's intelligence as AI systems accumulate domain expertise. The companies that start as autopilots will compound this advantage faster than anyone trying to transition from copilot to autopilot.

The six-to-one ratio says it all: for every dollar spent on software, six go to services.

That's where the next trillion-dollar company is hiding.

This is exactly why we rebuilt our consulting approach at alldone.consulting around AI-native frameworks. If you're launching products in this new landscape, DM me — happy to share our playbook for navigating the intelligence-to-judgment spectrum.